Source author record

Hong Liu

Hong Liu appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Catalog footprint

What is connected

133works

41topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2026arXiv

MoPO: Incorporating Motion Prior for Occluded Human Mesh Recovery

Although recent studies have made remarkable progress in human mesh recovery, they still exhibit limited robustness to occlusions and often produce inaccurate poses and severe motion jitter due to the insufficient spatial features for occluded body parts. Inspired by the rapid advancements in human motion prediction, we discover that compared to occluded image features, pose sequence inherently contains reliable motion prior for estimating occluded body parts. In this paper, we incorporate Motion Prior for Occluded human mesh recovery, called MoPO. Our MoPO mainly consists of two components: 1) The motion de-occlusion module, where we propose a spatial-temporal occlusion detector to detect joint visibility, and then we propose a lightweight motion predictor to complete the occluded body parts by predicting the most plausible joint positions based on history poses. 2) The motion-aware fusion and refinement module, which fuses the completed joint sequence with image features to estimate human shape and initial human pose. Moreover, the completed joint sequence is further used to refine the final human pose through inverse kinematics, which provides the occlusion-free motion prior for regressing human poses. Extensive experiments demonstrate that MoPO achieves state-of-the-art performance on both occlusion-specific and standard benchmarks, significantly enhancing the accuracy and temporal consistency of occluded human mesh recovery. Our code and demo can be found in the supplementary material.

preprint2023arXiv

A Gated Cross-domain Collaborative Network for Underwater Object Detection

Underwater object detection (UOD) plays a significant role in aquaculture and marine environmental protection. Considering the challenges posed by low contrast and low-light conditions in underwater environments, several underwater image enhancement (UIE) methods have been proposed to improve the quality of underwater images. However, only using the enhanced images does not improve the performance of UOD, since it may unavoidably remove or alter critical patterns and details of underwater objects. In contrast, we believe that exploring the complementary information from the two domains is beneficial for UOD. The raw image preserves the natural characteristics of the scene and texture information of the objects, while the enhanced image improves the visibility of underwater objects. Based on this perspective, we propose a Gated Cross-domain Collaborative Network (GCC-Net) to address the challenges of poor visibility and low contrast in underwater environments, which comprises three dedicated components. Firstly, a real-time UIE method is employed to generate enhanced images, which can improve the visibility of objects in low-contrast areas. Secondly, a cross-domain feature interaction module is introduced to facilitate the interaction and mine complementary information between raw and enhanced image features. Thirdly, to prevent the contamination of unreliable generated results, a gated feature fusion module is proposed to adaptively control the fusion ratio of cross-domain information. Our method presents a new UOD paradigm from the perspective of cross-domain information interaction and fusion. Experimental results demonstrate that the proposed GCC-Net achieves state-of-the-art performance on four underwater datasets.

preprint2023arXiv

A systematic formulation of chiral anomalous magnetohydrodynamics

We present a new way of deriving effective theories of dynamical electromagnetic fields in general media. It can be used to give a systematic formulation of magnetohydrodynamics (MHD) with strong magnetic fields, including systems with chiral matter and Adler-Bell-Jackiw (ABJ) anomaly. We work in the regime in which velocity and temperature fluctuations can be neglected. The resulting chiral anomalous MHD incorporates and generalizes the chiral magnetic effect, the chiral separation effect, the chiral electric separation effect, as well as recently derived strong-field MHD, all in a single coherent framework. At linearized level, the theory predicts that the chiral magnetic wave does not survive dynamical electromagnetic fields. A different chiral wave, to which we refer as the chiral magnetic electric separation wave, emerges as a result of dynamical versions of the chiral electric separation effect and the chiral magnetic effect. We predict its wave velocity. We also introduce a simple, but solvable nonlinear model to explore the fate of the chiral instability.

preprint2023arXiv

Achieving Domain Generalization in Underwater Object Detection by Domain Mixup and Contrastive Learning

The performance of existing underwater object detection methods degrades seriously when facing domain shift caused by complicated underwater environments. Due to the limitation of the number of domains in the dataset, deep detectors easily memorize a few seen domains, which leads to low generalization ability. There are two common ideas to improve the domain generalization performance. First, it can be inferred that the detector trained on as many domains as possible is domain-invariant. Second, for the images with the same semantic content in different domains, their hidden features should be equivalent. This paper further excavates these two ideas and proposes a domain generalization framework (named DMC) that learns how to generalize across domains from Domain Mixup and Contrastive Learning. First, based on the formation of underwater images, an image in an underwater environment is the linear transformation of another underwater environment. Thus, a style transfer model, which outputs a linear transformation matrix instead of the whole image, is proposed to transform images from one source domain to another, enriching the domain diversity of the training data. Second, mixup operation interpolates different domains on the feature level, sampling new domains on the domain manifold. Third, contrastive loss is selectively applied to features from different domains to force the model to learn domain invariant features but retain the discriminative capacity. With our method, detectors will be robust to domain shift. Also, a domain generalization benchmark S-UODAC2020 for detection is set up to measure the performance of our method. Comprehensive experiments on S-UODAC2020 and two object recognition benchmarks (PACS and VLCS) demonstrate that the proposed method is able to learn domain-invariant representations, and outperforms other domain generalization methods.

preprint2023arXiv

Many Hamiltonian subsets in large graphs with given density

A set of vertices in a graph is a Hamiltonian subset if it induces a subgraph containing a Hamiltonian cycle. Kim, Liu, Sharifzadeh and Staden proved that among all graphs with minimum degree $d$, $K_{d+1}$ minimises the number of Hamiltonian subsets. We prove a near optimal lower bound that takes also the order and the structure of a graph into account. For many natural graph classes, it provides a much better bound than the extremal one ($\approx 2^{d+1}$). Among others, our bound implies that an $n$-vertex $C_4$-free graphs with minimum degree $d$ contains at least $n2^{d^{2-o(1)}}$ Hamiltonian subsets.

preprint2023arXiv

Polynomial Schur's theorem

We resolve the Ramsey problem for $\{x,y,z:x+y=p(z)\}$ for all polynomials $p$ over $\mathbb{Z}$. In particular, we characterise all polynomials that are $2$-Ramsey, that is, those $p(z)$ such that any $2$-colouring of $\mathbb{N}$ contains infinitely many monochromatic solutions for $x+y=p(z)$. For polynomials that are not $2$-Ramsey, we characterise all $2$-colourings of $\mathbb{N}$ that are not $2$-Ramsey, revealing that certain divisibility barrier is the only obstruction to $2$-Ramseyness for $x+y=p(z)$.

preprint2022arXiv

Advancing Semi-Supervised Task Oriented Dialog Systems by JSA Learning of Discrete Latent Variable Models

Developing semi-supervised task-oriented dialog (TOD) systems by leveraging unlabeled dialog data has attracted increasing interests. For semi-supervised learning of latent state TOD models, variational learning is often used, but suffers from the annoying high-variance of the gradients propagated through discrete latent variables and the drawback of indirectly optimizing the target log-likelihood. Recently, an alternative algorithm, called joint stochastic approximation (JSA), has emerged for learning discrete latent variable models with impressive performances. In this paper, we propose to apply JSA to semi-supervised learning of the latent state TOD models, which is referred to as JSA-TOD. To our knowledge, JSA-TOD represents the first work in developing JSA based semi-supervised learning of discrete latent variable conditional models for such long sequential generation problems like in TOD systems. Extensive experiments show that JSA-TOD significantly outperforms its variational learning counterpart. Remarkably, semi-supervised JSA-TOD using 20% labels performs close to the full-supervised baseline on MultiWOZ2.1.

preprint2022arXiv

AO2-DETR: Arbitrary-Oriented Object Detection Transformer

Arbitrary-oriented object detection (AOOD) is a challenging task to detect objects in the wild with arbitrary orientations and cluttered arrangements. Existing approaches are mainly based on anchor-based boxes or dense points, which rely on complicated hand-designed processing steps and inductive bias, such as anchor generation, transformation, and non-maximum suppression reasoning. Recently, the emerging transformer-based approaches view object detection as a direct set prediction problem that effectively removes the need for hand-designed components and inductive biases. In this paper, we propose an Arbitrary-Oriented Object DEtection TRansformer framework, termed AO2-DETR, which comprises three dedicated components. More precisely, an oriented proposal generation mechanism is proposed to explicitly generate oriented proposals, which provides better positional priors for pooling features to modulate the cross-attention in the transformer decoder. An adaptive oriented proposal refinement module is introduced to extract rotation-invariant region features and eliminate the misalignment between region features and objects. And a rotation-aware set matching loss is used to ensure the one-to-one matching process for direct set prediction without duplicate predictions. Our method considerably simplifies the overall pipeline and presents a new AOOD paradigm. Comprehensive experiments on several challenging datasets show that our method achieves superior performance on the AOOD task.

preprint2022arXiv

Breast Cancer Molecular Subtypes Prediction on Pathological Images with Discriminative Patch Selecting and Multi-Instance Learning

Molecular subtypes of breast cancer are important references to personalized clinical treatment. For cost and labor savings, only one of the patient's paraffin blocks is usually selected for subsequent immunohistochemistry (IHC) to obtain molecular subtypes. Inevitable sampling error is risky due to tumor heterogeneity and could result in a delay in treatment. Molecular subtype prediction from conventional H&E pathological whole slide images (WSI) using AI method is useful and critical to assist pathologists pre-screen proper paraffin block for IHC. It's a challenging task since only WSI level labels of molecular subtypes can be obtained from IHC. Gigapixel WSIs are divided into a huge number of patches to be computationally feasible for deep learning. While with coarse slide-level labels, patch-based methods may suffer from abundant noise patches, such as folds, overstained regions, or non-tumor tissues. A weakly supervised learning framework based on discriminative patch selecting and multi-instance learning was proposed for breast cancer molecular subtype prediction from H&E WSIs. Firstly, co-teaching strategy was adopted to learn molecular subtype representations and filter out noise patches. Then, a balanced sampling strategy was used to handle the imbalance in subtypes in the dataset. In addition, a noise patch filtering algorithm that used local outlier factor based on cluster centers was proposed to further select discriminative patches. Finally, a loss function integrating patch with slide constraint information was used to finetune MIL framework on obtained discriminative patches and further improve the performance of molecular subtyping. The experimental results confirmed the effectiveness of the proposed method and our models outperformed even senior pathologists, with potential to assist pathologists to pre-screen paraffin blocks for IHC in clinic.

preprint2022arXiv

Clique immersion in graphs without fixed bipartite graph

A graph $G$ contains $H$ as an \emph{immersion} if there is an injective mapping $ϕ: V(H)\rightarrow V(G)$ such that for each edge $uv\in E(H)$, there is a path $P_{uv}$ in $G$ joining vertices $ϕ(u)$ and $ϕ(v)$, and all the paths $P_{uv}$, $uv\in E(H)$, are pairwise edge-disjoint. An analogue of Hadwiger's conjecture for the clique immersions by Lescure and Meyniel, and independently by Abu-Khzam and Langston, states that every graph $G$ contains $K_{χ(G)}$ as an immersion. We prove that for any constant $\varepsilon>0$ and integers $s,t\ge2$, there exists $d_0=d_0(\varepsilon,s,t)$ such that every $K_{s,t}$-free graph $G$ with $d(G)\ge d_0$ contains a clique immersion of order $(1-\varepsilon)d(G)$. This implies that the above-mentioned conjecture is asymptotically true for graphs without a fixed complete bipartite graph.

preprint2022arXiv

Contrastive Learning from Spatio-Temporal Mixed Skeleton Sequences for Self-Supervised Skeleton-Based Action Recognition

Self-supervised skeleton-based action recognition with contrastive learning has attracted much attention. Recent literature shows that data augmentation and large sets of contrastive pairs are crucial in learning such representations. In this paper, we found that directly extending contrastive pairs based on normal augmentations brings limited returns in terms of performance, because the contribution of contrastive pairs from the normal data augmentation to the loss get smaller as training progresses. Therefore, we delve into hard contrastive pairs for contrastive learning. Motivated by the success of mixing augmentation strategy which improves the performance of many tasks by synthesizing novel samples, we propose SkeleMixCLR: a contrastive learning framework with a spatio-temporal skeleton mixing augmentation (SkeleMix) to complement current contrastive learning approaches by providing hard contrastive samples. First, SkeleMix utilizes the topological information of skeleton data to mix two skeleton sequences by randomly combing the cropped skeleton fragments (the trimmed view) with the remaining skeleton sequences (the truncated view). Second, a spatio-temporal mask pooling is applied to separate these two views at the feature level. Third, we extend contrastive pairs with these two views. SkeleMixCLR leverages the trimmed and truncated views to provide abundant hard contrastive pairs since they involve some context information from each other due to the graph convolution operations, which allows the model to learn better motion representations for action recognition. Extensive experiments on NTU-RGB+D, NTU120-RGB+D, and PKU-MMD datasets show that SkeleMixCLR achieves state-of-the-art performance. Codes are available at https://github.com/czhaneva/SkeleMixCLR.

preprint2022arXiv

Disjoint isomorphic balanced clique subdivisions

A thoroughly studied problem in Extremal Graph Theory is to find the best possible density condition in a host graph $G$ for guaranteeing the presence of a particular subgraph $H$ in $G$. One such classical result, due to Bollobás and Thomason, and independently Komlós and Szemerédi, states that average degree $O(k^2)$ guarantees the existence of a $K_k$-subdivision. We study two directions extending this result. On the one hand, Verstraëte conjectured that the quadratic bound $O(k^2)$ would guarantee already two vertex-disjoint isomorphic copies of a $K_k$-subdivision. On the other hand, Thomassen conjectured that for each $k \in \mathbb{N}$ there is some $d = d(k)$ such that every graph with average degree at least $d$ contains a balanced subdivision of $K_k$, that is, a copy of $K_k$ where the edges are replaced by paths of equal length. Recently, Liu and Montgomery confirmed Thomassen's conjecture, but the optimal bound on $d(k)$ remains open. In this paper, we show that the quadratic bound $O(k^2)$ suffices to force a balanced $K_k$-subdivision. This gives the optimal bound on $d(k)$ needed in Thomassen's conjecture and implies the existence of $O(1)$ many vertex-disjoint isomorphic $K_k$-subdivisions, confirming Verstraëte's conjecture in a strong sense.

preprint2022arXiv

Exploiting Temporal Contexts with Strided Transformer for 3D Human Pose Estimation

Despite the great progress in 3D human pose estimation from videos, it is still an open problem to take full advantage of a redundant 2D pose sequence to learn representative representations for generating one 3D pose. To this end, we propose an improved Transformer-based architecture, called Strided Transformer, which simply and effectively lifts a long sequence of 2D joint locations to a single 3D pose. Specifically, a Vanilla Transformer Encoder (VTE) is adopted to model long-range dependencies of 2D pose sequences. To reduce the redundancy of the sequence, fully-connected layers in the feed-forward network of VTE are replaced with strided convolutions to progressively shrink the sequence length and aggregate information from local contexts. The modified VTE is termed as Strided Transformer Encoder (STE), which is built upon the outputs of VTE. STE not only effectively aggregates long-range information to a single-vector representation in a hierarchical global and local fashion, but also significantly reduces the computation cost. Furthermore, a full-to-single supervision scheme is designed at both full sequence and single target frame scales applied to the outputs of VTE and STE, respectively. This scheme imposes extra temporal smoothness constraints in conjunction with the single target frame supervision and hence helps produce smoother and more accurate 3D poses. The proposed Strided Transformer is evaluated on two challenging benchmark datasets, Human3.6M and HumanEva-I, and achieves state-of-the-art results with fewer parameters. Code and models are available at \url{https://github.com/Vegetebird/StridedTransformer-Pose3D}.

preprint2022arXiv

Exponential decay of intersection volume with applications on list-decodability and Gilbert-Varshamov type bound

We give some natural sufficient conditions for balls in a metric space to have small intersection. Roughly speaking, this happens when the metric space is (i) expanding and (ii) well-spread, and (iii) a certain random variable on the boundary of a ball has a small tail. As applications, we show that the volume of intersection of balls in Hamming, Johnson spaces and symmetric groups decay exponentially as their centers drift apart. To verify condition (iii), we prove some large deviation inequalities `on a slice' for functions with Lipschitz conditions. We then use these estimates on intersection volumes to $\bullet$ obtain a sharp lower bound on list-decodability of random $q$-ary codes, confirming a conjecture of Li and Wootters; and $\bullet$ improve the classical bound of Levenshtein from 1971 on constant weight codes by a factor linear in dimension, resolving a problem raised by Jiang and Vardy. Our probabilistic point of view also offers a unified framework to obtain improvements on other Gilbert--Varshamov type bounds, giving conceptually simple and calculation-free proofs for $q$-ary codes, permutation codes, and spherical codes. Another consequence is a counting result on the number of codes, showing ampleness of large codes.

preprint2022arXiv

Facial-Sketch Synthesis: A New Challenge

This paper aims to conduct a comprehensive study on facial-sketch synthesis (FSS). However, due to the high costs of obtaining hand-drawn sketch datasets, there lacks a complete benchmark for assessing the development of FSS algorithms over the last decade. We first introduce a high-quality dataset for FSS, named FS2K, which consists of 2,104 image-sketch pairs spanning three types of sketch styles, image backgrounds, lighting conditions, skin colors, and facial attributes. FS2K differs from previous FSS datasets in difficulty, diversity, and scalability and should thus facilitate the progress of FSS research. Second, we present the largest-scale FSS investigation by reviewing 89 classical methods, including 25 handcrafted feature-based facial-sketch synthesis approaches, 29 general translation methods, and 35 image-to-sketch approaches. Besides, we elaborate comprehensive experiments on the existing 19 cutting-edge models. Third, we present a simple baseline for FSS, named FSGAN. With only two straightforward components, i.e., facial-aware masking and style-vector expansion, FSGAN surpasses the performance of all previous state-of-the-art models on the proposed FS2K dataset by a large margin. Finally, we conclude with lessons learned over the past years and point out several unsolved challenges. Our code is available at https://github.com/DengPingFan/FSGAN.

preprint2022arXiv

High-order Photonic Cavity Modes Enabled 3D Structural Color

It remains a challenge to directly print three-dimensional arbitrary shapes that exhibit structural colors at the micrometer scale. Woodpile photonic crystals (WPCs) fabricated via two-photon lithography (TPL) are promising as building blocks to produce 3D geometries that generate structural colors due to their ability to exhibit either omnidirectional or anisotropic photonic stopbands. However, existing approaches have focused on achieving structural colors when illuminating WPCs from the top, which necessitates print resolutions beyond the limit of commercial TPL and/or post-processing techniques. Here, we devised a new strategy to support high-order photonic cavity modes upon side-illumination on WPCs that surprisingly generate large reflectance peaks in the visible spectrum. Based on that, we demonstrate one-step printing of 3D photonic structural colors without requiring post-processing or subwavelength features. Vivid colors with reflectance peaks exhibiting a full width at half maximum of ~25 nm, a maximum reflectance of 50%, gamut of ~85% of sRGB, and large viewing angles, were achieved. In addition, we also demonstrated voxel-level manipulation and control of colors in arbitrary-shaped 3D objects constituted with WPCs as unit cells, which has great potential for applications in dynamic color displays, colorimetric sensing, anti-counterfeiting, and light-matter interaction platforms.

preprint2022arXiv

How to build a pillar: a proof of Thomassen's conjecture

Carsten Thomassen in 1989 conjectured that if a graph has minimum degree more than the number of atoms in the universe ($δ(G)\ge 10^{10^{10}}$), then it contains a pillar, which is a graph that consists of two vertex-disjoint cycles of the same length, $s$ say, along with $s$ vertex-disjoint paths of the same length which connect matching vertices in order around the cycles. Despite the simplicity of the structure of pillars and various developments of powerful embedding methods for paths and cycles in the past three decades, this innocent looking conjecture has seen no progress to date. In this paper, we give a proof of this conjecture by building a pillar (algorithmically) in sublinear expanders.

preprint2022arXiv

Identity-Sensitive Knowledge Propagation for Cloth-Changing Person Re-identification

Cloth-changing person re-identification (CC-ReID), which aims to match person identities under clothing changes, is a new rising research topic in recent years. However, typical biometrics-based CC-ReID methods often require cumbersome pose or body part estimators to learn cloth-irrelevant features from human biometric traits, which comes with high computational costs. Besides, the performance is significantly limited due to the resolution degradation of surveillance images. To address the above limitations, we propose an effective Identity-Sensitive Knowledge Propagation framework (DeSKPro) for CC-ReID. Specifically, a Cloth-irrelevant Spatial Attention module is introduced to eliminate the distraction of clothing appearance by acquiring knowledge from the human parsing module. To mitigate the resolution degradation issue and mine identity-sensitive cues from human faces, we propose to restore the missing facial details using prior facial knowledge, which is then propagated to a smaller network. After training, the extra computations for human parsing or face restoration are no longer required. Extensive experiments show that our framework outperforms state-of-the-art methods by a large margin. Our code is available at https://github.com/KimbingNg/DeskPro.

preprint2022arXiv

Learning Deep Direct-Path Relative Transfer Function for Binaural Sound Source Localization

Direct-path relative transfer function (DP-RTF) refers to the ratio between the direct-path acoustic transfer functions of two microphone channels. Though DP-RTF fully encodes the sound spatial cues and serves as a reliable localization feature, it is often erroneously estimated in the presence of noise and reverberation. This paper proposes to learn DP-RTF with deep neural networks for robust binaural sound source localization. A DP-RTF learning network is designed to regress the binaural sensor signals to a real-valued representation of DP-RTF. It consists of a branched convolutional neural network module to separately extract the inter-channel magnitude and phase patterns, and a convolutional recurrent neural network module for joint feature learning. To better explore the speech spectra to aid the DP-RTF estimation, a monaural speech enhancement network is used to recover the direct-path spectrograms from the noisy ones. The enhanced spectrograms are stacked onto the noisy spectrograms to act as the input of the DP-RTF learning network. We train one unique DP-RTF learning network using many different binaural arrays to enable the generalization of DP-RTF learning across arrays. This way avoids time-consuming training data collection and network retraining for a new array, which is very useful in practical application. Experimental results on both simulated and real-world data show the effectiveness of the proposed method for direction of arrival (DOA) estimation in the noisy and reverberant environment, and a good generalization ability to unseen binaural arrays.

preprint2022arXiv

MHFormer: Multi-Hypothesis Transformer for 3D Human Pose Estimation

Estimating 3D human poses from monocular videos is a challenging task due to depth ambiguity and self-occlusion. Most existing works attempt to solve both issues by exploiting spatial and temporal relationships. However, those works ignore the fact that it is an inverse problem where multiple feasible solutions (i.e., hypotheses) exist. To relieve this limitation, we propose a Multi-Hypothesis Transformer (MHFormer) that learns spatio-temporal representations of multiple plausible pose hypotheses. In order to effectively model multi-hypothesis dependencies and build strong relationships across hypothesis features, the task is decomposed into three stages: (i) Generate multiple initial hypothesis representations; (ii) Model self-hypothesis communication, merge multiple hypotheses into a single converged representation and then partition it into several diverged hypotheses; (iii) Learn cross-hypothesis communication and aggregate the multi-hypothesis features to synthesize the final 3D pose. Through the above processes, the final representation is enhanced and the synthesized pose is much more accurate. Extensive experiments show that MHFormer achieves state-of-the-art results on two challenging datasets: Human3.6M and MPI-INF-3DHP. Without bells and whistles, its performance surpasses the previous best result by a large margin of 3% on Human3.6M. Code and models are available at \url{https://github.com/Vegetebird/MHFormer}.

preprint2022arXiv

Mission Apollo: Landing Optical Circuit Switching at Datacenter Scale

In this paper, we describe Apollo, to the best of our knowledge, the world's first large-scale production deployment of optical circuit switches (OCSes) for datacenter networking. We will first describe the infrastructure challenges and use cases that motivated optical switching inside datacenters. We then delve into the requirements of OCSes for datacenter applications: balancing cost, port count, switching time, and optical performance, which drive design choices and implementation details of our internally developed 3D MEMS-based OCS. To enable the Apollo optical switching layer, we employ circulators to realize bidirectional links through the OCS, effectively doubling the OCS radix. The OCS and circulator design choices were critical for meeting network bandwidth, scale, and cost targets. We review the critical co-design of WDM transceiver technology for these OCS plus circulator-based bidirectional links and their corresponding physical impairments, delivered over four generations/speeds of optical interconnect. Finally, we conclude with thoughts on future directions in hardware development and associated applications.

preprint2022arXiv

Multi-Scale Spatial Temporal Graph Convolutional Network for Skeleton-Based Action Recognition

Graph convolutional networks have been widely used for skeleton-based action recognition due to their excellent modeling ability of non-Euclidean data. As the graph convolution is a local operation, it can only utilize the short-range joint dependencies and short-term trajectory but fails to directly model the distant joints relations and long-range temporal information that are vital to distinguishing various actions. To solve this problem, we present a multi-scale spatial graph convolution (MS-GC) module and a multi-scale temporal graph convolution (MT-GC) module to enrich the receptive field of the model in spatial and temporal dimensions. Concretely, the MS-GC and MT-GC modules decompose the corresponding local graph convolution into a set of sub-graph convolution, forming a hierarchical residual architecture. Without introducing additional parameters, the features will be processed with a series of sub-graph convolutions, and each node could complete multiple spatial and temporal aggregations with its neighborhoods. The final equivalent receptive field is accordingly enlarged, which is capable of capturing both short- and long-range dependencies in spatial and temporal domains. By coupling these two modules as a basic block, we further propose a multi-scale spatial temporal graph convolutional network (MST-GCN), which stacks multiple blocks to learn effective motion representations for action recognition. The proposed MST-GCN achieves remarkable performance on three challenging benchmark datasets, NTU RGB+D, NTU-120 RGB+D and Kinetics-Skeleton, for skeleton-based action recognition.

preprint2022arXiv

PetLock:A Genderless and Standard Interface for the Future On-orbit Construction

Modular design is the foundation of on orbit construction technology of large space facilities in the future.Standard interface is the key technology of modular design of the future space robotic systems and space facilities.This paper presents the designed and tested of PetLock,a standard and genderless interface which can transfer mechanical loads,power and data between the future modular space robotic manipulator and spacecraft.PetLock adopts a completely genderless design,including connection face,locking mechanism,data and power interface.The connection surface provides a large translation and rotation misalignment tolerance,due to its 120-degree symmetrical and 3D shape design.The locking mechanism features the three locking pins retraction structure design,which is simple and reliable.POGO pin connectors in the center of the interface provides the power and data transfer capabilities.Due to the advantages of high locking force,large tolerance,high reliability and low cost,PetLock has the very big application potential in future on orbit construction missions.

preprint2022arXiv

Prototype Design and Efficiency Analysis of a Novel Robot Drive Based on 3K-H-V Topology

Robot actuators directly affect the performance of robots, and robot drives directly affect the performance of robot actuators. With the development of robotics, robots have put higher requirements on robot drives, such as high stiffness, high accuracy, high loading, high efficiency, low backlash, compact size, and hollow structure. In order to meet the demand development of robot actuators, this research base proposes a new robot drive based on 3K-H-V topology using involute and cycloidal gear shapes, planetary cycloidal drive, from the perspective of drive topology and through the design idea of decoupling. In this study, the reduction ratio and the efficiency model of the 3K-H-V topology were analyzed, and a prototype planetary cycloidal actuator was designed. The feasibility of the drive is initially verified by experimentally concluding that the PCA has a hollow structure, compact size, and high torque density (69 kg/Nm).

preprint2022arXiv

Recent advances and clinical applications of deep learning in medical image analysis

Deep learning has received extensive research interest in developing new medical image processing algorithms, and deep learning based models have been remarkably successful in a variety of medical imaging tasks to support disease detection and diagnosis. Despite the success, the further improvement of deep learning models in medical image analysis is majorly bottlenecked by the lack of large-sized and well-annotated datasets. In the past five years, many studies have focused on addressing this challenge. In this paper, we reviewed and summarized these recent studies to provide a comprehensive overview of applying deep learning methods in various medical image analysis tasks. Especially, we emphasize the latest progress and contributions of state-of-the-art unsupervised and semi-supervised deep learning in medical image analysis, which are summarized based on different application scenarios, including classification, segmentation, detection, and image registration. We also discuss the major technical challenges and suggest the possible solutions in future research efforts.

preprint2022arXiv

Recent Advances in Tunable Metasurfaces: Materials, Design and Applications

Metasurfaces, a two-dimensional (2D) form of metamaterials constituted by planar meta-atoms, exhibit exotic abilities to freely tailor electromagnetic (EM) waves. Over the past decade, tunable metasurfaces have come to the frontier in the field of nanophotonics, with tremendous effort focused on developing and integrating various active materials into metasurfaces. As a result, tunable/reconfigurable metasurfaces with multi-functionalities triggered by various external stimuli have been successfully demonstrated, openings a new avenue to dynamically manipulate and control EM waves for photonic applications in demand. In this review, we first brief the progress of tunable metasurfaces development in the last decade and highlight representative works from the perspectives of active materials development, design methodologies and application-driven exploration. Then, we elaborate on the active tuning mechanisms and relevant active materials. Next, we discuss recent achievements in theory as well as machine learning (ML) assisted design methodologies to sustain the development of this field. After that, we summarize and describe typical application areas of the tunable metasurfaces. We conclude this review by analyzing existing challenges and presenting our perspectives on future directions and opportunities in this vibrant and fast-developing field.

preprint2022arXiv

Self-supervised Learning is More Robust to Dataset Imbalance

Self-supervised learning (SSL) is a scalable way to learn general visual representations since it learns without labels. However, large-scale unlabeled datasets in the wild often have long-tailed label distributions, where we know little about the behavior of SSL. In this work, we systematically investigate self-supervised learning under dataset imbalance. First, we find out via extensive experiments that off-the-shelf self-supervised representations are already more robust to class imbalance than supervised representations. The performance gap between balanced and imbalanced pre-training with SSL is significantly smaller than the gap with supervised learning, across sample sizes, for both in-domain and, especially, out-of-domain evaluation. Second, towards understanding the robustness of SSL, we hypothesize that SSL learns richer features from frequent data: it may learn label-irrelevant-but-transferable features that help classify the rare classes and downstream tasks. In contrast, supervised learning has no incentive to learn features irrelevant to the labels from frequent examples. We validate this hypothesis with semi-synthetic experiments and theoretical analyses on a simplified setting. Third, inspired by the theoretical insights, we devise a re-weighted regularization technique that consistently improves the SSL representation quality on imbalanced datasets with several evaluation criteria, closing the small gap between balanced and imbalanced datasets with the same number of examples.

preprint2022arXiv

Shape of the asymptotic maximum sum-free sets in integer lattice grids

We determine the shape of all sum-free sets in $\{1,2,\ldots,n\}^2$ of size close to the maximum $\frac{3}{5}n^2$, solving a problem of Elsholtz and Rackham. We show that all such asymptotic maximum sum-free sets lie completely in the stripe $\frac{4}{5}n-o(n)\le x+y\le\frac{8}{5}n+ o(n)$. We also determine for any positive integer $p$ the maximum size of a subset $A\subseteq \{1,2,\ldots,n\}^2$ which forbids the triple $(x,y,z)$ satisfying $px+py=z$.

preprint2022arXiv

Simultaneous Alignment and Surface Regression Using Hybrid 2D-3D Networks for 3D Coherent Layer Segmentation of Retina OCT Images

Automated surface segmentation of retinal layer is important and challenging in analyzing optical coherence tomography (OCT). Recently, many deep learning based methods have been developed for this task and yield remarkable performance. However, due to large spatial gap and potential mismatch between the B-scans of OCT data, all of them are based on 2D segmentation of individual B-scans, which may loss the continuity information across the B-scans. In addition, 3D surface of the retina layers can provide more diagnostic information, which is crucial in quantitative image analysis. In this study, a novel framework based on hybrid 2D-3D convolutional neural networks (CNNs) is proposed to obtain continuous 3D retinal layer surfaces from OCT. The 2D features of individual B-scans are extracted by an encoder consisting of 2D convolutions. These 2D features are then used to produce the alignment displacement field and layer segmentation by two 3D decoders, which are coupled via a spatial transformer module. The entire framework is trained end-to-end. To the best of our knowledge, this is the first study that attempts 3D retinal layer segmentation in volumetric OCT images based on CNNs. Experiments on a publicly available dataset show that our framework achieves superior results to state-of-the-art 2D methods in terms of both layer segmentation accuracy and cross-B-scan 3D continuity, thus offering more clinical values than previous works.

preprint2022arXiv

Snowmass White Paper: Effective Field Theories for Condensed Matter Systems

We review recent progress and a number of future directions for applications of effective field theory methods to condensed matter systems broadly defined. Our emphasis is on areas that have allowed a fertile exchange of ideas between high energy physics and many-body theory. We discuss developments in the effective field theory of spontaneous symmetry breaking, of hydrodynamics and non-equilibrium dynamics more generally, fracton phases of matter, and dualities between 2+1 dimensional field theories. We furthermore discuss the application of effective field theory to non-Fermi liquids, the dynamics of entanglement entropy and to condensed matter aspects of cosmology.

preprint2022arXiv

Snowmass White Paper: New ideas for many-body quantum systems from string theory and black holes

During the last two decades many new insights into the dynamics of strongly coupled quantum many-body systems have been obtained using gauge/gravity duality, with black holes often playing a universal role. In this white paper we summarize the results obtained and offer some outlook for future developments, including the ongoing mutually beneficial feedback loop with the study of more general, not necessarily holographic, quantum many-body systems.

preprint2022arXiv

SP-SEDT: Self-supervised Pre-training for Sound Event Detection Transformer

Recently, an event-based end-to-end model (SEDT) has been proposed for sound event detection (SED) and achieves competitive performance. However, compared with the frame-based model, it requires more training data with temporal annotations to improve the localization ability. Synthetic data is an alternative, but it suffers from a great domain gap with real recordings. Inspired by the great success of UP-DETR in object detection, we propose to self-supervisedly pre-train SEDT (SP-SEDT) by detecting random patches (only cropped along the time axis). Experiments on the DCASE2019 task4 dataset show the proposed SP-SEDT can outperform fine-tuned frame-based model. The ablation study is also conducted to investigate the impact of different loss functions and patch size.

preprint2022arXiv

Spatiotemporal Propagation Learning for Network-Wide Flight Delay Prediction

Demystifying the delay propagation mechanisms among multiple airports is fundamental to precise and interpretable delay prediction, which is crucial during decision-making for all aviation industry stakeholders. The principal challenge lies in effectively leveraging the spatiotemporal dependencies and exogenous factors related to the delay propagation. However, previous works only consider limited spatiotemporal patterns with few factors. To promote more comprehensive propagation modeling for delay prediction, we propose SpatioTemporal Propagation Network (STPN), a space-time separable graph convolutional network, which is novel in spatiotemporal dependency capturing. From the aspect of spatial relation modeling, we propose a multi-graph convolution model considering both geographic proximity and airline schedule. From the aspect of temporal dependency capturing, we propose a multi-head self-attentional mechanism that can be learned end-to-end and explicitly reason multiple kinds of temporal dependency of delay time series. We show that the joint spatial and temporal learning models yield a sum of the Kronecker product, which factors the spatiotemporal dependence into the sum of several spatial and temporal adjacency matrices. By this means, STPN allows cross-talk of spatial and temporal factors for modeling delay propagation. Furthermore, a squeeze and excitation module is added to each layer of STPN to boost meaningful spatiotemporal features. To this end, we apply STPN to multi-step ahead arrival and departure delay prediction in large-scale airport networks. To validate the effectiveness of our model, we experiment with two real-world delay datasets, including U.S and China flight delays; and we show that STPN outperforms state-of-the-art methods. In addition, counterfactuals produced by STPN show that it learns explainable delay propagation patterns.

preprint2022arXiv

SRP-DNN: Learning Direct-Path Phase Difference for Multiple Moving Sound Source Localization

Multiple moving sound source localization in real-world scenarios remains a challenging issue due to interaction between sources, time-varying trajectories, distorted spatial cues, etc. In this work, we propose to use deep learning techniques to learn competing and time-varying direct-path phase differences for localizing multiple moving sound sources. A causal convolutional recurrent neural network is designed to extract the direct-path phase difference sequence from signals of each microphone pair. To avoid the assignment ambiguity and the problem of uncertain output-dimension encountered when simultaneously predicting multiple targets, the learning target is designed in a weighted sum format, which encodes source activity in the weight and direct-path phase differences in the summed value. The learned direct-path phase differences for all microphone pairs can be directly used to construct the spatial spectrum according to the formulation of steered response power (SRP). This deep neural network (DNN) based SRP method is referred to as SRP-DNN. The locations of sources are estimated by iteratively detecting and removing the dominant source from the spatial spectrum, in which way the interaction between sources is reduced. Experimental results on both simulated and real-world data show the superiority of the proposed method in the presence of noise and reverberation.

preprint2022arXiv

Transformer for Single Image Super-Resolution

Single image super-resolution (SISR) has witnessed great strides with the development of deep learning. However, most existing studies focus on building more complex networks with a massive number of layers. Recently, more and more researchers start to explore the application of Transformer in computer vision tasks. However, the heavy computational cost and high GPU memory occupation of the vision Transformer cannot be ignored. In this paper, we propose a novel Efficient Super-Resolution Transformer (ESRT) for SISR. ESRT is a hybrid model, which consists of a Lightweight CNN Backbone (LCB) and a Lightweight Transformer Backbone (LTB). Among them, LCB can dynamically adjust the size of the feature map to extract deep features with a low computational cost. LTB is composed of a series of Efficient Transformers (ET), which occupies a small GPU memory occupation, thanks to the specially designed Efficient Multi-Head Attention (EMHA). Extensive experiments show that ESRT achieves competitive results with low computational costs. Compared with the original Transformer which occupies 16,057M GPU memory, ESRT only occupies 4,191M GPU memory. All codes are available at https://github.com/luissen/ESRT.

preprint2022arXiv

Transformers Improve Breast Cancer Diagnosis from Unregistered Multi-View Mammograms

Deep convolutional neural networks (CNNs) have been widely used in various medical imaging tasks. However, due to the intrinsic locality of convolution operation, CNNs generally cannot model long-range dependencies well, which are important for accurately identifying or mapping corresponding breast lesion features computed from unregistered multiple mammograms. This motivates us to leverage the architecture of Multi-view Vision Transformers to capture long-range relationships of multiple mammograms from the same patient in one examination. For this purpose, we employ local Transformer blocks to separately learn patch relationships within four mammograms acquired from two-view (CC/MLO) of two-side (right/left) breasts. The outputs from different views and sides are concatenated and fed into global Transformer blocks, to jointly learn patch relationships between four images representing two different views of the left and right breasts. To evaluate the proposed model, we retrospectively assembled a dataset involving 949 sets of mammograms, which include 470 malignant cases and 479 normal or benign cases. We trained and evaluated the model using a five-fold cross-validation method. Without any arduous preprocessing steps (e.g., optimal window cropping, chest wall or pectoral muscle removal, two-view image registration, etc.), our four-image (two-view-two-side) Transformer-based model achieves case classification performance with an area under ROC curve (AUC = 0.818), which significantly outperforms AUC = 0.784 achieved by the state-of-the-art multi-view CNNs (p = 0.009). It also outperforms two one-view-two-side models that achieve AUC of 0.724 (CC view) and 0.769 (MLO view), respectively. The study demonstrates the potential of using Transformers to develop high-performing computer-aided diagnosis schemes that combine four mammograms.

preprint2022arXiv

Virtual Adversarial Training for Semi-supervised Breast Mass Classification

This study aims to develop a novel computer-aided diagnosis (CAD) scheme for mammographic breast mass classification using semi-supervised learning. Although supervised deep learning has achieved huge success across various medical image analysis tasks, its success relies on large amounts of high-quality annotations, which can be challenging to acquire in practice. To overcome this limitation, we propose employing a semi-supervised method, i.e., virtual adversarial training (VAT), to leverage and learn useful information underlying in unlabeled data for better classification of breast masses. Accordingly, our VAT-based models have two types of losses, namely supervised and virtual adversarial losses. The former loss acts as in supervised classification, while the latter loss aims at enhancing model robustness against virtual adversarial perturbation, thus improving model generalizability. To evaluate the performance of our VAT-based CAD scheme, we retrospectively assembled a total of 1024 breast mass images, with equal number of benign and malignant masses. A large CNN and a small CNN were used in this investigation, and both were trained with and without the adversarial loss. When the labeled ratios were 40% and 80%, VAT-based CNNs delivered the highest classification accuracy of 0.740 and 0.760, respectively. The experimental results suggest that the VAT-based CAD scheme can effectively utilize meaningful knowledge from unlabeled data to better classify mammographic breast mass images.

preprint2022arXiv

Weakly-supervised 3D Human Pose Estimation with Cross-view U-shaped Graph Convolutional Network

Although monocular 3D human pose estimation methods have made significant progress, it is far from being solved due to the inherent depth ambiguity. Instead, exploiting multi-view information is a practical way to achieve absolute 3D human pose estimation. In this paper, we propose a simple yet effective pipeline for weakly-supervised cross-view 3D human pose estimation. By only using two camera views, our method can achieve state-of-the-art performance in a weakly-supervised manner, requiring no 3D ground truth but only 2D annotations. Specifically, our method contains two steps: triangulation and refinement. First, given the 2D keypoints that can be obtained through any classic 2D detection methods, triangulation is performed across two views to lift the 2D keypoints into coarse 3D poses. Then, a novel cross-view U-shaped graph convolutional network (CV-UGCN), which can explore the spatial configurations and cross-view correlations, is designed to refine the coarse 3D poses. In particular, the refinement progress is achieved through weakly-supervised learning, in which geometric and structure-aware consistency checks are performed. We evaluate our method on the standard benchmark dataset, Human3.6M. The Mean Per Joint Position Error on the benchmark dataset is 27.4 mm, which outperforms existing state-of-the-art methods remarkably (27.4 mm vs 30.2 mm).

preprint2021arXiv

Bound entanglement in thermalized states and black hole radiation

We study the mixed-state entanglement structure of chaotic quantum many-body systems at late times using the recently developed $\textit{equilibrium approximation}$. A rich entanglement phase diagram emerges when we generalize this technique to evaluate the logarithmic negativity for various universality classes of macroscopically thermalized states. Unlike in the infinite temperature case, when we impose energy constraints at finite temperature, the phase diagrams for the logarithmic negativity and the mutual information become distinct. In particular, we identify a regime where the negativity is extensive but the mutual information is sub-extensive, indicating a large amount of $\textit{bound entanglement}$. When applied to evaporating black holes, these results imply that there is quantum entanglement within the Hawking radiation long before the Page time, although this entanglement may not be distillable into EPR pairs.

preprint2021arXiv

Mixed-state entanglement and information recovery in thermalized states and evaporating black holes

We study the universal behavior of quantum information-theoretic quantities in thermalized isolated quantum many-body systems and evaporating black holes. In particular, we study a genuine mixed-state entanglement measure called the logarithmic negativity, other correlation measures including the Renyi negativities and the mutual information, and a signature of multipartite entanglement called the reflected entropy. We also probe the feasibility of recovering quantum information from subsystems of a thermalized quantum many-body system or from the radiation of an evaporating black hole, using quantities such as relative entropy and Petz map fidelity. A recently developed technique called the equilibrium approximation allows us to probe these quantities at finite temperature. We find striking qualitative differences from the infinite temperature case, which has been the topic of previous studies using Haar-random states. In particular, we find regimes where the logarithmic negativity is extensive but the mutual information is sub-extensive, indicating a large amount of undistillable, bound entanglement in thermalized states. For evaporating black holes at finite temperature, both the logarithmic negativity and the Petz map fidelity reveal an important new time scale $t_b$, which is earlier than the Page time $t_p$ by a finite fraction of the total evaporation time. We find that $t_b$, as opposed to $t_p$, is the time scale at which quantum entanglement between different parts of the radiation becomes extensive, and the fidelity of information recovery for a large diary thrown into the black hole starts to grow.

preprint2021arXiv

On systems of maximal quantum chaos

A remarkable feature of chaos in many-body quantum systems is the existence of a bound on the quantum Lyapunov exponent. An important question is to understand what is special about maximally chaotic systems which saturate this bound. Here we provide further evidence for the `hydrodynamic' origin of chaos in such systems, and discuss hallmarks of maximally chaotic systems. We first provide evidence that a hydrodynamic effective field theory of chaos we previously proposed should be understood as a theory of maximally chaotic systems. We then emphasize and make explicit a signature of maximal chaos which was only implicit in prior literature, namely the suppression of exponential growth in commutator squares of generic few-body operators. We provide a general argument for this suppression within our chaos effective field theory, and illustrate it using SYK models and holographic systems. We speculate that this suppression indicates that the nature of operator scrambling in maximally chaotic systems is fundamentally different to scrambling in non-maximally chaotic systems. We also discuss a simplest scenario for the existence of a maximally chaotic regime at sufficiently large distances even for non-maximally chaotic systems.

preprint2020arXiv

Anti-Bandit Neural Architecture Search for Model Defense

Deep convolutional neural networks (DCNNs) have dominated as the best performers in machine learning, but can be challenged by adversarial attacks. In this paper, we defend against adversarial attacks using neural architecture search (NAS) which is based on a comprehensive search of denoising blocks, weight-free operations, Gabor filters and convolutions. The resulting anti-bandit NAS (ABanditNAS) incorporates a new operation evaluation measure and search process based on the lower and upper confidence bounds (LCB and UCB). Unlike the conventional bandit algorithm using UCB for evaluation only, we use UCB to abandon arms for search efficiency and LCB for a fair competition between arms. Extensive experiments demonstrate that ABanditNAS is faster than other NAS methods, while achieving an $8.73\%$ improvement over prior arts on CIFAR-10 under PGD-$7$.

preprint2020arXiv

Applying a random projection algorithm to optimize machine learning model for breast lesion classification

Machine learning is widely used in developing computer-aided diagnosis (CAD) schemes of medical images. However, CAD usually computes large number of image features from the targeted regions, which creates a challenge of how to identify a small and optimal feature vector to build robust machine learning models. In this study, we investigate feasibility of applying a random projection algorithm to build an optimal feature vector from the initially CAD-generated large feature pool and improve performance of machine learning model. We assemble a retrospective dataset involving 1,487 cases of mammograms in which 644 cases have confirmed malignant mass lesions and 843 have benign lesions. A CAD scheme is first applied to segment mass regions and initially compute 181 features. Then, support vector machine (SVM) models embedded with several feature dimensionality reduction methods are built to predict likelihood of lesions being malignant. All SVM models are trained and tested using a leave-one-case-out cross-validation method. SVM generates a likelihood score of each segmented mass region depicting on one-view mammogram. By fusion of two scores of the same mass depicting on two-view mammograms, a case-based likelihood score is also evaluated. Comparing with the principle component analyses, nonnegative matrix factorization, and Chi-squared methods, SVM embedded with the random projection algorithm yielded a significantly higher case-based lesion classification performance with the area under ROC curve of 0.84+0.01 (p<0.02). The study demonstrates that the random project algorithm is a promising method to generate optimal feature vectors to help improve performance of machine learning models of medical images.

preprint2020arXiv

Bi-directional Exponential Angular Triplet Loss for RGB-Infrared Person Re-Identification

RGB-Infrared person re-identification (RGB-IR Re- ID) is a cross-modality matching problem, where the modality discrepancy is a big challenge. Most existing works use Euclidean metric based constraints to resolve the discrepancy between features of images from different modalities. However, these methods are incapable of learning angularly discriminative feature embedding because Euclidean distance cannot measure the included angle between embedding vectors effectively. As an angularly discriminative feature space is important for classifying the human images based on their embedding vectors, in this paper, we propose a novel ranking loss function, named Bi-directional Exponential Angular Triplet Loss, to help learn an angularly separable common feature space by explicitly constraining the included angles between embedding vectors. Moreover, to help stabilize and learn the magnitudes of embedding vectors, we adopt a common space batch normalization layer. The quantitative and qualitative experiments on the SYSU-MM01 and RegDB dataset support our analysis. On SYSU-MM01 dataset, the performance is improved from 7.40% / 11.46% to 38.57% / 38.61% for rank-1 accuracy / mAP compared with the baseline. The proposed method can be generalized to the task of single-modality Re-ID and improves the rank-1 accuracy / mAP from 92.0% / 81.7% to 94.7% / 86.6% on the Market-1501 dataset, from 82.6% / 70.6% to 87.6% / 77.1% on the DukeMTMC-reID dataset. Code: https://github.com/prismformore/expAT

preprint2020arXiv

Dynamical Phase Transition from Nonequilibrium Dynamics of Dark Solitons

By holographic duality, we identify a novel dynamical phase transition which results from the temperature dependence of non-equilibrium dynamics of dark solitons in a superfluid.For a non-equilibrium superfluid system with an initial density of dark solitons, there exists a critical temperature $T_d$,above which the system relaxes to equilibrium by producing sound waves, while below which it goes through an intermediate phase with a finite density of vortex-antivortex pairs. In particular, as $T_d$ is approached from below, the density of vortex pairs scales as $(T_d - T)^γ$ with the critical exponent $γ= 1/2$.

preprint2020arXiv

Guided learning for weakly-labeled semi-supervised sound event detection

We propose a simple but efficient method termed Guided Learning for weakly-labeled semi-supervised sound event detection (SED). There are two sub-targets implied in weakly-labeled SED: audio tagging and boundary detection. Instead of designing a single model by considering a trade-off between the two sub-targets, we design a teacher model aiming at audio tagging to guide a student model aiming at boundary detection to learn using the unlabeled data. The guidance is guaranteed by the audio tagging performance gap of the two models. In the meantime, the student model liberated from the trade-off is able to provide more excellent boundary detection results. We propose a principle to design such two models based on the relation between the temporal compression scale and the two sub-targets. We also propose an end-to-end semi-supervised learning process for these two models to enable their abilities to rise alternately. Experiments on the DCASE2018 Task4 dataset show that our approach achieves competitive performance.

preprint2020arXiv

Hadamard Matrix Guided Online Hashing

Online image hashing has attracted increasing research attention recently, which receives large-scale data in a streaming manner to update the hash functions on-the-fly. Its key challenge lies in the difficulty of balancing the learning timeliness and model accuracy. To this end, most works follow a supervised setting, i.e., using class labels to boost the hashing performance, which defects in two aspects: First, strong constraints, e.g., orthogonal or similarity preserving, are used, which however are typically relaxed and lead to large accuracy drop. Second, large amounts of training batches are required to learn the up-to-date hash functions, which largely increase the learning complexity. To handle the above challenges, a novel supervised online hashing scheme termed Hadamard Matrix Guided Online Hashing (HMOH) is proposed in this paper. Our key innovation lies in introducing Hadamard matrix, which is an orthogonal binary matrix built via Sylvester method. In particular, to release the need of strong constraints, we regard each column of Hadamard matrix as the target code for each class label, which by nature satisfies several desired properties of hashing codes. To accelerate the online training, LSH is first adopted to align the lengths of target code and to-be-learned binary code. We then treat the learning of hash functions as a set of binary classification problems to fit the assigned target code. Finally, extensive experiments demonstrate the superior accuracy and efficiency of the proposed method over various state-of-the-art methods. Codes are available at https://github.com/lmbxmu/mycode.

preprint2020arXiv

Multi-Branch Learning for Weakly-Labeled Sound Event Detection

There are two sub-tasks implied in the weakly-supervised SED: audio tagging and event boundary detection. Current methods which combine multi-task learning with SED requires annotations both for these two sub-tasks. Since there are only annotations for audio tagging available in weakly-supervised SED, we design multiple branches with different learning purposes instead of pursuing multiple tasks. Similar to multiple tasks, multiple different learning purposes can also prevent the common feature which the multiple branches share from overfitting to any one of the learning purposes. We design these multiple different learning purposes based on combinations of different MIL strategies and different pooling methods. Experiments on the DCASE 2018 Task 4 dataset and the URBAN-SED dataset both show that our method achieves competitive performance.

preprint2020arXiv

Online Initialization and Extrinsic Spatial-Temporal Calibration for Monocular Visual-Inertial Odometry

This paper presents an online initialization method for bootstrapping the optimization-based monocular visual-inertial odometry (VIO). The method can online calibrate the relative transformation (spatial) and time offsets (temporal) among camera and IMU, as well as estimate the initial values of metric scale, velocity, gravity, gyroscope bias, and accelerometer bias during the initialization stage. To compensate for the impact of time offset, our method includes two short-term motion interpolation algorithms for the camera and IMU pose estimation. Besides, it includes a three-step process to incrementally estimate the parameters from coarse to fine. First, the extrinsic rotation, gyroscope bias, and time offset are estimated by minimizing the rotation difference between the camera and IMU. Second, the metric scale, gravity, and extrinsic translation are approximately estimated by using the compensated camera poses and ignoring the accelerometer bias. Third, these values are refined by taking into account the accelerometer bias and the gravitational magnitude. For further optimizing the system states, a nonlinear optimization algorithm, which considers the time offset, is introduced for global and local optimization. Experimental results on public datasets show that the initial values and the extrinsic parameters, as well as the sensor poses, can be accurately estimated by the proposed method.

preprint2020arXiv

Projection & Probability-Driven Black-Box Attack

Generating adversarial examples in a black-box setting retains a significant challenge with vast practical application prospects. In particular, existing black-box attacks suffer from the need for excessive queries, as it is non-trivial to find an appropriate direction to optimize in the high-dimensional space. In this paper, we propose Projection & Probability-driven Black-box Attack (PPBA) to tackle this problem by reducing the solution space and providing better optimization. For reducing the solution space, we first model the adversarial perturbation optimization problem as a process of recovering frequency-sparse perturbations with compressed sensing, under the setting that random noise in the low-frequency space is more likely to be adversarial. We then propose a simple method to construct a low-frequency constrained sensing matrix, which works as a plug-and-play projection matrix to reduce the dimensionality. Such a sensing matrix is shown to be flexible enough to be integrated into existing methods like NES and Bandits$_{TD}$. For better optimization, we perform a random walk with a probability-driven strategy, which utilizes all queries over the whole progress to make full use of the sensing matrix for a less query budget. Extensive experiments show that our method requires at most 24% fewer queries with a higher attack success rate compared with state-of-the-art approaches. Finally, the attack method is evaluated on the real-world online service, i.e., Google Cloud Vision API, which further demonstrates our practical potentials.

preprint2020arXiv

Quantum many-body physics from a gravitational lens

The last two decades have seen the emergence of stunning interconnections among various previously remotely related disciplines such as condensed matter, nuclear physics, gravity and quantum information, fueled both by experimental advances and new powerful theoretical methods brought by holographic duality. In this non-technical review we sample some recent developments in holographic duality in connection with quantum many-body dynamics. These include insights into strongly correlated phases without quasiparticles and their transport properties, quantum many-body chaos, and scrambling of quantum information. We also discuss recent progress in understanding the structure of holographic duality itself using quantum information, including a "local" version of the duality as well as the quantum error correction interpretation of quantum many-body states with a gravity dual, and how such notions help demonstrate the unitarity of black hole evaporation.

preprint2020arXiv

Self-Refining Deep Symmetry Enhanced Network for Rain Removal

Rain removal aims to remove the rain streaks on rain images. The state-of-the-art methods are mostly based on Convolutional Neural Network~(CNN). However, as CNN is not equivariant to object rotation, these methods are unsuitable for dealing with the tilted rain streaks. To tackle this problem, we propose Deep Symmetry Enhanced Network~(DSEN) that is able to explicitly extract the rotation equivariant features from rain images. In addition, we design a self-refining mechanism to remove the accumulated rain streaks in a coarse-to-fine manner. This mechanism reuses DSEN with a novel information link which passes the gradient flow to the higher stages. Extensive experiments on both synthetic and real-world rain images show that our self-refining DSEN yields the top performance.

preprint2020arXiv

Spatial Pyramid Based Graph Reasoning for Semantic Segmentation

The convolution operation suffers from a limited receptive filed, while global modeling is fundamental to dense prediction tasks, such as semantic segmentation. In this paper, we apply graph convolution into the semantic segmentation task and propose an improved Laplacian. The graph reasoning is directly performed in the original feature space organized as a spatial pyramid. Different from existing methods, our Laplacian is data-dependent and we introduce an attention diagonal matrix to learn a better distance metric. It gets rid of projecting and re-projecting processes, which makes our proposed method a light-weight module that can be easily plugged into current computer vision architectures. More importantly, performing graph reasoning directly in the feature space retains spatial relationships and makes spatial pyramid possible to explore multiple long-range contextual patterns from different scales. Experiments on Cityscapes, COCO Stuff, PASCAL Context and PASCAL VOC demonstrate the effectiveness of our proposed methods on semantic segmentation. We achieve comparable performance with advantages in computational and memory overhead.

preprint2020arXiv

Specialized Decision Surface and Disentangled Feature for Weakly-Supervised Polyphonic Sound Event Detection

In this paper, a special decision surface for the weakly-supervised sound event detection (SED) and a disentangled feature (DF) for the multi-label problem in polyphonic SED are proposed. We approach SED as a multiple instance learning (MIL) problem and utilize a neural network framework with a pooling module to solve it. General MIL approaches include two kinds: the instance-level approaches and embedding-level approaches. We present a method of generating instance-level probabilities for the embedding level approaches which tend to perform better than the instance-level approaches in terms of bag-level classification but can not provide instance-level probabilities in current approaches. Moreover, we further propose a specialized decision surface (SDS) for the embedding-level attention pooling. We analyze and explained why an embedding-level attention module with SDS is better than other typical pooling modules from the perspective of the high-level feature space. As for the problem of the unbalanced dataset and the co-occurrence of multiple categories in the polyphonic event detection task, we propose a DF to reduce interference among categories, which optimizes the high-level feature space by disentangling it based on class-wise identifiable information and obtaining multiple different subspaces. Experiments on the dataset of DCASE 2018 Task 4 show that the proposed SDS and DF significantly improve the detection performance of the embedding-level MIL approach with an attention pooling module and outperform the first place system in the challenge by 6.6 percentage points.

preprint2020arXiv

The exact minimum number of triangles in graphs of given order and size

What is the minimum number of triangles in a graph of given order and size? Motivated by earlier results of Mantel and Turán, Rademacher solved the first non-trivial case of this problem in 1941. The problem was revived by Erdős in 1955; it is now known as the Erdős-Rademacher problem. After attracting much attention, it was solved asymptotically in a major breakthrough by Razborov in 2008. In this paper, we provide an exact solution for all large graphs whose edge density is bounded away from~$1$, which in this range confirms a conjecture of Lovász and Simonovits from 1975. Furthermore, we give a description of the extremal graphs.

preprint2020arXiv

Video Logo Retrieval based on local Features

Estimation of the frequency and duration of logos in videos is important and challenging in the advertisement industry as a way of estimating the impact of ad purchases. Since logos occupy only a small area in the videos, the popular methods of image retrieval could fail. This paper develops an algorithm called Video Logo Retrieval (VLR), which is an image-to-video retrieval algorithm based on the spatial distribution of local image descriptors that measure the distance between the query image (the logo) and a collection of video images. VLR uses local features to overcome the weakness of global feature-based models such as convolutional neural networks (CNN). Meanwhile, VLR is flexible and does not require training after setting some hyper-parameters. The performance of VLR is evaluated on two challenging open benchmark tasks (SoccerNet and Standford I2V), and compared with other state-of-the-art logo retrieval or detection algorithms. Overall, VLR shows significantly higher accuracy compared with the existing methods.

preprint2020arXiv

When Dictionary Learning Meets Deep Learning: Deep Dictionary Learning and Coding Network for Image Recognition with Limited Data

We present a new Deep Dictionary Learning and Coding Network (DDLCN) for image recognition tasks with limited data. The proposed DDLCN has most of the standard deep learning layers (e.g., input/output, pooling, fully connected, etc.), but the fundamental convolutional layers are replaced by our proposed compound dictionary learning and coding layers. The dictionary learning learns an over-complete dictionary for input training data. At the deep coding layer, a locality constraint is added to guarantee that the activated dictionary bases are close to each other. Then the activated dictionary atoms are assembled and passed to the compound dictionary learning and coding layers. In this way, the activated atoms in the first layer can be represented by the deeper atoms in the second dictionary. Intuitively, the second dictionary is designed to learn the fine-grained components shared among the input dictionary atoms, thus a more informative and discriminative low-level representation of the dictionary atoms can be obtained. We empirically compare DDLCN with several leading dictionary learning methods and deep learning models. Experimental results on five popular datasets show that DDLCN achieves competitive results compared with state-of-the-art methods when the training data is limited. Code is available at https://github.com/Ha0Tang/DDLCN.

preprint2020arXiv

WQT and DG-YOLO: towards domain generalization in underwater object detection

A General Underwater Object Detector (GUOD) should perform well on most of underwater circumstances. However, with limited underwater dataset, conventional object detection methods suffer from domain shift severely. This paper aims to build a GUOD with small underwater dataset with limited types of water quality. First, we propose a data augmentation method Water Quality Transfer (WQT) to increase domain diversity of the original small dataset. Second, for mining the semantic information from data generated by WQT, DG-YOLO is proposed, which consists of three parts: YOLOv3, DIM and IRM penalty. Finally, experiments on original and synthetic URPC2019 dataset prove that WQT+DG-YOLO achieves promising performance of domain generalization in underwater object detection.

preprint2019arXiv

Compact all-fiber polarization-independent up-conversion single-photon detector

We demonstrate a compact all-fiber polarization-independent up-conversion single-photon detector based on integrated reverse proton exchanged periodically poled lithium niobate waveguides. The horizontally and vertically polarized components of randomly polarized signals are separated with a fiber-coupled polarization beam splitter, launched into two orthogonally polarized polarization maintaining fibers and fetched into two adjacent independent waveguides on the same device. The up-converted outputs from both waveguide channels are then combined with a multi-mode fiber combiner and detected by a silicon detector. With this configuration, the polarization-independent single-photon counting at 1.55 um is achieved with a system detection efficiency of 29.3%, a dark count rate of 1600 counts per second, and a polarization dependent loss of 0.1dB. This compact all-fiber system is robust and has great application potential in practical quantum key distribution systems.

preprint2019arXiv

Sample phase gradient and fringe phase shift in dual phase grating X-ray interferometry

One of the key tasks in grating based x-ray phase contrast imaging is to accurately retrieve local phase gradients of a sample from measured intensity fringe shifts. To fulfill this task in dual phase grating interferometry, one needs to know the exact mathematical relationship between the two. In this work, using intuitive analysis of the sample-generated fringe shifts based on the beat pattern formation mechanism, the authors derived the formulas relating sample's phase gradients to fringe phase shifts. These formulas provide also a design optimization tool for dual phase grating interferometry.

preprint2019arXiv

Tungsten Boride: a 2D Multiple Dirac Semimetal for Hydrogen Evolution Reaction

Here, we propose a two-dimensional tungsten boride (WB4) lattice, with the Gibbs free energy for the adsorption of atomic hydrogen, tending to be the ideal value of 0 eV at 3% strained state, to host a better hydrogen evolution reaction activity. Based on first-principles calculations, it is demonstrated that the multiple d-p-pi and d-p-sigma Dirac conjugations of WB4 lattice ensures its excellent electronic transport characteristics. Meanwhile, coupling with the d-orbitals of W, the p-orbitals of borophene subunits in WB4 lattice can modulate the d band center to get a good HER performance. Our results not only provide a versatile platform for hosting multiple Dirac semimetal states with a sandwich configuration, but also offer a guiding principle for discovering the relationship between intrinsic properties of the active centre and the catalytic activity of metal layer from the emerging field of low-dimensional noble-metal-free lattices.

preprint2016arXiv

Coulomb drag in topological insulator films

We study Coulomb drag between the top and bottom surfaces of topological insulator films. We derive a kinetic equation for the thin-film spin density matrix containing the full spin structure of the two-layer system, and analyze the electron-electron interaction in detail in order to recover all terms responsible for Coulomb drag. Focusing on typical topological insulator systems, with film thicknesses d up to 6 nm, we obtain numerical and approximate analytical results for the drag resistivity $ρ_\text{D}$ and find that $ρ_\text{D}$ is proportional to $T^2d^{-4}n^{-3/2}_{\text{a}}n^{-3/2}_{\text{p}}$ at low temperature T and low electron density $n_{\text{a,p}}$, with a denoting the active layer and p the passive layer. In addition, we compare $ρ_{\text{D}}$ with graphene, identifying qualitative and quantitative differences, and we discuss the multi valley case, ultra thin films and electron-hole layers.

preprint2016arXiv

Densities in large permutations and parameter testing

A classical theorem of Erdos, Lovasz and Spencer asserts that the densities of connected subgraphs in large graphs are independent. We prove an analogue of this theorem for permutations and we then apply the methods used in the proof to give an example of a finitely approximable permutation parameter that is not finitely forcible. The latter answers a question posed by two of the authors and Moreira and Sampaio.

preprint2016arXiv

Flat Helical Nanosieves

Compact and miniaturized devices with flexible functionalities are always highly demanded in optical integrated systems. Plasmonic nanosieve has been successfully harnessed as an ultrathin flat platform for complex manipulation of light, including holography, vortex generation and non-linear processes. Compared with most of reported single-functional devices, multi-functional nanosieves might find more complex and novel applications across nano-photonics, optics and nanotechnology. Here, we experimentally demonstrate a promising roadmap for nanosieve-based helical devices, which achieves full manipulations of optical vortices, including its generation, hybridization, spatial multiplexing, focusing and non-diffraction propagation etc., by controlling the geometric phase of spin light via over 121 thousands of spatially-rotated nano-sieves. Thanks to such spin-conversion nanosieve helical elements, it is no longer necessary to employ the conventional two-beam interferometric measurement to characterize optical vortices, while the interference can be realized natively without changing any parts of the current setup. The proposed strategy makes the far-field manipulations of optical orbital angular momentum within an ultrathin interface viable and bridges singular optics and integrated optics. In addition, it enables more unique extensibility and flexibility in versatile optical elements than traditional phase-accumulated helical optical devices.

preprint2016arXiv

Local conditions for exponentially many subdivisions

Given a graph $F$, let $s_t(F)$ be the number of subdivisions of $F$, each with a different vertex set, which one can guarantee in a graph $G$ in which every edge lies in at least $t$ copies of $F$. In 1990, Tuza asked for which graphs $F$ and large $t$, one has that $s_t(F)$ is exponential in a power of $t$. We show that, somewhat surprisingly, the only such $F$ are complete graphs, and for every $F$ which is not complete, $s_t(F)$ is polynomial in $t$. Further, for a natural strengthening of the local condition above, we also characterise those $F$ for which $s_t(F)$ is exponential in a power of $t$.

preprint2016arXiv

Ordinal Constrained Binary Code Learning for Nearest Neighbor Search

Recent years have witnessed extensive attention in binary code learning, a.k.a. hashing, for nearest neighbor search problems. It has been seen that high-dimensional data points can be quantized into binary codes to give an efficient similarity approximation via Hamming distance. Among existing schemes, ranking-based hashing is recent promising that targets at preserving ordinal relations of ranking in the Hamming space to minimize retrieval loss. However, the size of the ranking tuples, which shows the ordinal relations, is quadratic or cubic to the size of training samples. By given a large-scale training data set, it is very expensive to embed such ranking tuples in binary code learning. Besides, it remains a dificulty to build ranking tuples efficiently for most ranking-preserving hashing, which are deployed over an ordinal graph-based setting. To handle these problems, we propose a novel ranking-preserving hashing method, dubbed Ordinal Constraint Hashing (OCH), which efficiently learns the optimal hashing functions with a graph-based approximation to embed the ordinal relations. The core idea is to reduce the size of ordinal graph with ordinal constraint projection, which preserves the ordinal relations through a small data set (such as clusters or random samples). In particular, to learn such hash functions effectively, we further relax the discrete constraints and design a specific stochastic gradient decent algorithm for optimization. Experimental results on three large-scale visual search benchmark datasets, i.e. LabelMe, Tiny100K and GIST1M, show that the proposed OCH method can achieve superior performance over the state-of-the-arts approaches.

preprint2016arXiv

Orientation Driven Bag of Appearances for Person Re-identification

Person re-identification (re-id) consists of associating individual across camera network, which is valuable for intelligent video surveillance and has drawn wide attention. Although person re-identification research is making progress, it still faces some challenges such as varying poses, illumination and viewpoints. For feature representation in re-identification, existing works usually use low-level descriptors which do not take full advantage of body structure information, resulting in low representation ability. %discrimination. To solve this problem, this paper proposes the mid-level body-structure based feature representation (BSFR) which introduces body structure pyramid for codebook learning and feature pooling in the vertical direction of human body. Besides, varying viewpoints in the horizontal direction of human body usually causes the data missing problem, $i.e.$, the appearances obtained in different orientations of the identical person could vary significantly. To address this problem, the orientation driven bag of appearances (ODBoA) is proposed to utilize person orientation information extracted by orientation estimation technic. To properly evaluate the proposed approach, we introduce a new re-identification dataset (Market-1203) based on the Market-1501 dataset and propose a new re-identification dataset (PKU-Reid). Both datasets contain multiple images captured in different body orientations for each person. Experimental results on three public datasets and two proposed datasets demonstrate the superiority of the proposed approach, indicating the effectiveness of body structure and orientation information for improving re-identification performance.

preprint2016arXiv

Spread of entanglement and causality

We investigate causality constraints on the time evolution of entanglement entropy after a global quench in relativistic theories. We first provide a general proof that the so-called tsunami velocity is bounded by the speed of light. We then generalize the free particle streaming model of arXiv:cond-mat/0503393 to general dimensions and to an arbitrary entanglement pattern of the initial state. In more than two spacetime dimensions the spread of entanglement in these models is highly sensitive to the initial entanglement pattern, but we are able to prove an upper bound on the normalized rate of growth of entanglement entropy, and hence the tsunami velocity. The bound is smaller than what one gets for quenches in holographic theories, which highlights the importance of interactions in the spread of entanglement in many-body systems. We propose an interacting model which we believe provides an upper bound on the spread of entanglement for interacting relativistic theories. In two spacetime dimensions with multiple intervals, this model and its variations are able to reproduce intricate results exhibited by holographic theories for a significant part of the parameter space. For higher dimensions, the model bounds the tsunami velocity at the speed of light. Finally, we construct a geometric model for entanglement propagation based on a tensor network construction for global quenches.

preprint2016arXiv

Supervised Matrix Factorization for Cross-Modality Hashing

Matrix factorization has been recently utilized for the task of multi-modal hashing for cross-modality visual search, where basis functions are learned to map data from different modalities to the same Hamming embedding. In this paper, we propose a novel cross-modality hashing algorithm termed Supervised Matrix Factorization Hashing (SMFH) which tackles the multi-modal hashing problem with a collective non-matrix factorization across the different modalities. In particular, SMFH employs a well-designed binary code learning algorithm to preserve the similarities among multi-modal original features through a graph regularization. At the same time, semantic labels, when available, are incorporated into the learning procedure. We conjecture that all these would facilitate to preserve the most relevant information during the binary quantization process, and hence improve the retrieval accuracy. We demonstrate the superior performance of SMFH on three cross-modality visual search benchmarks, i.e., the PASCAL-Sentence, Wiki, and NUS-WIDE, with quantitative comparison to various state-of-the-art methods

preprint2016arXiv

The number of subsets of integers with no $k$-term arithmetic progression

Addressing a question of Cameron and Erd\Ho s, we show that, for infinitely many values of $n$, the number of subsets of $\{1,2,\ldots, n\}$ that do not contain a $k$-term arithmetic progression is at most $2^{O(r_k(n))}$, where $r_k(n)$ is the maximum cardinality of a subset of $\{1,2,\ldots, n\}$ without a $k$-term arithmetic progression. This bound is optimal up to a constant factor in the exponent. For all values of $n$, we prove a weaker bound, which is nevertheless sufficient to transfer the current best upper bound on $r_k(n)$ to the sparse random setting. To achieve these bounds, we establish a new supersaturation result, which roughly states that sets of size $Θ(r_k(n))$ contain superlinearly many $k$-term arithmetic progressions. For integers $r$ and $k$, Erd\Ho s asked whether there is a set of integers $S$ with no $(k+1)$-term arithmetic progression, but such that any $r$-coloring of $S$ yields a monochromatic $k$-term arithmetic progression. Nešetřil and Rödl, and independently Spencer, answered this question affirmatively. We show the following density version: for every $k\ge 3$ and $δ>0$, there exists a reasonably dense subset of primes $S$ with no $(k+1)$-term arithmetic progression, yet every $U\subseteq S$ of size $|U|\geδ|S|$ contains a $k$-term arithmetic progression. Our proof uses the hypergraph container method, which has proven to be a very powerful tool in extremal combinatorics. The idea behind the container method is to have a small certificate set to describe a large independent set. We give two further applications in the appendix using this idea.

preprint2015arXiv

Continuously Shaping Orbital Angular Momentum with an Analog Optical Vortex Transmitter

Dynamic generation of obitial angular momentum (OAM) of light has enabled complex manipulation of micro-particles, high-dimension quantum entanglement and optical communication. We report an analog vortex transmitter made of one bilaterally symmetric grating and an aperture, emitting optical vortices with the average OAM value continuously variant in the entire rational range. Benefiting from linearly-varying transverse dislocation along its axis of symmetry, this diffractive transmitter possesses extra degree of freedom in engineering broadband optical vortices meanwhile preserving a novel spiniform phase with equally spaced singularities. It unlimitedly increases the average OAM of light by embracing more singularities, which is significantly different from that for Laguerre-Gaussian (LG) and Bessel vortex beams. Realizing analog generation of OAM in a single device, this technique can be potentially extended to other frequencies and applied to a wide spectrum of developments on quantum physics, aperiodic photonics and optical manipulation.

preprint2015arXiv

Defect formation beyond Kibble-Zurek mechanism and holography

We study the dynamic after a smooth quench across a continuous transition from the disordered phase to the ordered phase. Based on scaling ideas, linear response and the spectrum of unstable modes, we develop a theoretical framework, valid for any second order phase transition, for the early-time evolution of the condensate in the broken phase. Our analysis unveils a novel period of non-adiabatic evolution after the system passes through the phase transition, where a parametrically large amount of coarsening occurs before a well-defined condensate forms. Our formalism predicts a rate of defect formation parametrically smaller than the Kibble-Zurek prediction and yields a criterion for the break-down of Kibble-Zurek scaling for sufficiently fast quenches. We numerically test our formalism for a thermal quench in a 2 + 1 dimensional holographic superfluid. These findings, of direct relevance in a broad range of fields including cold atom, condensed matter, statistical mechanism and cosmology, are an important step towards a more quantitative understanding of dynamical phase transitions.

preprint2015arXiv

Intersecting families of discrete structures are typically trivial

The study of intersecting structures is central to extremal combinatorics. A family of permutations $\mathcal{F} \subset S_n$ is \emph{$t$-intersecting} if any two permutations in $\mathcal{F}$ agree on some $t$ indices, and is \emph{trivial} if all permutations in $\mathcal{F}$ agree on the same $t$ indices. A $k$-uniform hypergraph is \emph{$t$-intersecting} if any two of its edges have $t$ vertices in common, and \emph{trivial} if all its edges share the same $t$ vertices. The fundamental problem is to determine how large an intersecting family can be. Ellis, Friedgut and Pilpel proved that for $n$ sufficiently large with respect to $t$, the largest $t$-intersecting families in $S_n$ are the trivial ones. The classic Erdős--Ko--Rado theorem shows that the largest $t$-intersecting $k$-uniform hypergraphs are also trivial when $n$ is large. We determine the \emph{typical} structure of $t$-intersecting families, extending these results to show that almost all intersecting families are trivial. We also obtain sparse analogues of these extremal results, showing that they hold in random settings. Our proofs use the Bollobás set-pairs inequality to bound the number of maximal intersecting families, which can then be combined with known stability theorems. We also obtain similar results for vector spaces.

preprint2015arXiv

Off-shell hydrodynamics from holography

We outline a program for obtaining an action principle for dissipative fluid dynamics by considering the holographic Wilsonian renormalization group applied to systems with a gravity dual. As a first step, in this paper we restrict to systems with a non-dissipative horizon. By integrating out gapped degrees of freedom in the bulk gravitational system between an asymptotic boundary and a horizon, we are led to a formulation of hydrodynamics where the dynamical variables are not standard velocity and temperature fields, but the relative embedding of the boundary and horizon hypersurfaces. At zeroth order, this action reduces to that proposed by Dubovsky et al. as an off-shell formulation of ideal fluid dynamics.

preprint2015arXiv

The typical structure of graphs with no large cliques

In 1987, Kolaitis, Prömel and Rothschild proved that, for every fixed $r \in \mathbb{N}$, almost every $n$-vertex $K_{r+1}$-free graph is $r$-partite. In this paper we extend this result to all functions $r = r(n)$ with $r \leqslant (\log n)^{1/4}$. The proof combines a new (close to sharp) supersaturation version of the Erdős-Simonovits stability theorem, the hypergraph container method, and a counting technique developed by Balogh, Bollobás and Simonovits.

preprint2015arXiv

The typical structure of maximal triangle-free graphs

Recently, settling a question of Erdős, Balogh and Petříčková showed that there are at most $2^{n^2/8+o(n^2)}$ $n$-vertex maximal triangle-free graphs, matching the previously known lower bound. Here we characterize the typical structure of maximal triangle-free graphs. We show that almost every maximal triangle-free graph $G$ admits a vertex partition $X\cup Y$ such that $G[X]$ is a perfect matching and $Y$ is an independent set. Our proof uses the Ruzsa-Szemerédi removal lemma, the Erdős-Simonovits stability theorem, and recent results of Balogh-Morris-Samotij and Saxton-Thomason on characterization of the structure of independent sets in hypergraphs. The proof also relies on a new bound on the number of maximal independent sets in triangle-free graphs with many vertex-disjoint $P_3$'s, which is of independent interest.

preprint2014arXiv

Angular Momentum Generation by Parity Violation

We generalize our holographic derivation of spontaneous angular momentum generation in 2 + 1 dimensions in several directions. We consider cases when a parity violating perturbation responsible for the angular momentum generation can be non-marginal (while in our previous paper we restricted to a marginal perturbation), including all possible two-derivative interactions, with parity violations triggered both by gauge and gravitational Chern-Simons terms in the bulk. We make only a minimal assumption about the bulk geometry that it is asymptotically AdS, respects the Poincaré symmetry in 2 + 1 dimensions, and has a horizon. In this generic setup, we find a remarkably concise and universal formula for the expectation value of the angular momentum density, to all orders in the parity violating perturbation.

preprint2014arXiv

Hall Viscosity and Angular Momentum in Gapless Holographic Models

We use the holographic approach to compare the Hall viscosity $η_H$ and the angular momentum density ${\cal J}$ in gapless systems in $2+1$ dimensions at finite temperature. We start with a conformal fixed point and turn on a perturbation which breaks the parity and time reversal symmetries via gauge and gravitational Chern-Simons couplings in the bulk. While the ratio of $η_H$ and ${\cal J}$ shows some universal properties when the perturbation is slightly relevant, we find that the two quantities behave differently in general. In particular, $η_H$ depends only on infrared physics, while ${\cal J}$ receives contributions from degrees of freedom at all scales.

preprint2014arXiv

On the number of $K_4$-saturating edges

Let $G$ be a $K_4$-free graph, an edge in its complement is a $K_4$-\emph{saturating} edge if the addition of this edge to $G$ creates a copy of $K_4$. Erdős and Tuza conjectured that for any $n$-vertex $K_4$-free graph $G$ with $\lfloor n^2/4\rfloor+1$ edges, one can find at least $(1+o(1))\frac{n^2}{16}$ $K_4$-saturating edges. We construct a graph with only $\frac{2n^2}{33}$ $K_4$-saturating edges. Furthermore, we prove that it is best possible, i.e., one can always find at least $(1+o(1))\frac{2n^2}{33}$ $K_4$-saturating edges in an $n$-vertex $K_4$-free graph with $\lfloor n^2/4\rfloor+1$ edges.

preprint2014arXiv

Quantum phase transitions in semi-local quantum liquids

We consider several types of quantum critical phenomena from finite-density gauge-gravity duality which to different degrees lie outside the Landau-Ginsburg-Wilson paradigm. These include: (1) a "bifurcating" critical point, for which the order parameter remains gapped at the critical point, and thus is not driven by soft order parameter fluctuations. Rather it appears to be driven by "confinement" which arises when two fixed points annihilate and lose conformality. On the condensed side, there is an infinite tower of condensed states and the nonlinear response of the tower exhibits an infinite spiral structure; (2) a "hybridized" critical point which can be described by a standard Landau-Ginsburg sector of order parameter fluctuations hybridized with a strongly coupled sector; (3) a "marginal" critical point which is obtained by tuning the above two critical points to occur together and whose bosonic fluctuation spectrum coincides with that postulated to underly the "Marginal Fermi Liquid" description of the optimally doped cuprates.

preprint2014arXiv

Screening, Friedel oscillations and low-temperature conductivity in topological insulator thin films

In thin topological insulator films, the top and bottom surfaces are coupled by tunneling, which restores backscattering and strongly affects screening. We calculate the dielectric function in the random phase approximation obtaining a closed-form result. Unlike independent TI surfaces, the dielectric function of thin films exhibits a valley as a function of wavenumber $q$ and tunneling, as well as a cusp at $q=2k_F$, with $k_F$ the Fermi wave vector. As a result of the cusp, Friedel oscillations decay with distance $r$ as $\sin(2k_Fr)/(2k_Fr)^2$. We determine the longitudinal conductivity $σ$ in the first Born approximation at low temperatures where screened impurities provide the dominant scattering mechanism. At high electron densities $n_e$, $σ\propto n_e$, while at low densities $σ\propto n_e^{3/2}$.

preprint2014arXiv

Subdivisions of a large clique in $C_6$-free graphs

Mader conjectured that every $C_4$-free graph has a subdivision of a clique of order linear in its average degree. We show that every $C_6$-free graph has such a subdivision of a large clique. We also prove the dense case of Mader's conjecture in a stronger sense, i.e. for every $c$, there is a $c'$ such that every $C_4$-free graph with average degree $cn^{1/2}$ has a subdivision of a clique $K_\ell$ with $\ell=\lfloor c'n^{1/2}\rfloor$ where every edge is subdivided exactly $3$ times.

preprint2014arXiv

Subwavelength focusing of azimuthally polarized beams with vortical phase in dielectrics by using an ultra-thin lens

We demonstrate that a planar and ultrathin binary lens can focus an azimuthally polarized beam with vortical phase (APV) to a subwavelength spot of transverse polarization. The results elaborates that, in the multi-layer medium, this focused spot, which is beyond the Rayleigh diffraction limitation, can be well maintained for several wavelengths after travelling through the dielectric interfaces, which is not attainable by using other vector beams (i.e., radially, linearly and circularly polarized beams) as the illuminating light. This compact optical system can be valuable in data writing and defect identification of wafer or silicon chips, owing to the enhanced polarized focusing through interfaces. It also enables to be highly integrated with traditional microscopy for the far-field super-resolution imaging, surface scanning and detection, and subwavelength focusing, owing to the enhanced focusing performance (reduced width and extended length) as well as the planarized configuration of the ultrathin lens.

preprint2014arXiv

The number of maximal sum-free subsets of integers

Cameron and Erdős raised the question of how many maximal sum-free sets there are in $\{1, \dots , n\}$, giving a lower bound of $2^{\lfloor n/4 \rfloor }$. In this paper we prove that there are in fact at most $2^{(1/4+o(1))n}$ maximal sum-free sets in $\{1, \dots , n\}$. Our proof makes use of container and removal lemmas of Green as well as a result of Deshouillers, Freiman, Sós and Temkin on the structure of sum-free sets.

preprint2013arXiv

A refinement of entanglement entropy and the number of degrees of freedom

We introduce a "renormalized entanglement entropy" which is intrinsically UV finite and is most sensitive to the degrees of freedom at the scale of the size R of the entangled region. We illustrated the power of this construction by showing that the qualitative behavior of the entanglement entropy for a non-Fermi liquid can be obtained by simple dimensional analysis. We argue that the functional dependence of the "renormalized entanglement entropy" on R can be interpreted as describing the renormalization group flow of the entanglement entropy with distance scale. The corresponding quantity for a spherical region in the vacuum, has some particularly interesting properties. For a conformal field theory, it reduces to the previously proposed central charge in all dimensions, and for a general quantum field theory, it interpolates between the central charges of the UV and IR fixed points as R is varied from zero to infinity. We conjecture that in three (spacetime) dimensions, it is always non-negative and monotonic, and provides a measure of the number of degrees of freedom of a system at scale R. In four dimensions, however, we find examples in which it is neither monotonic nor non-negative.

preprint2013arXiv

A Survey on Gas Sensing Technology

Sensing technology has been widely investigated and utilized for gas detection. Due to the different applicability and inherent limitations of different gas sensing technologies, researchers have been working on different scenarios with enhanced gas sensor calibration. This paper reviews the descriptions, evaluation, comparison and recent developments in existing gas sensing technologies. A classification of sensing technologies is given, based on the variation of electrical and other properties. Detailed introduction to sensing methods based on electrical variation is discussed through further classification according to sensing materials, including metal oxide semiconductors, polymers, carbon nanotubes, and moisture absorbing materials. Methods based on other kinds of variations such as optical, calorimetric, acoustic and gas-chromatographic, are presented in a general way. Several suggestions related to future development are also discussed. Furthermore, this paper focuses on sensitivity and selectivity for performance indicators to compare different sensing technologies, analyzes the factors that influence these two indicators, and lists several corresponding improved approaches.

preprint2013arXiv

Charge transport by holographic Fermi surfaces

We compute the contribution to the conductivity from holographic Fermi surfaces obtained from probe fermions in an AdS charged black hole. This requires calculating a certain part of the one-loop correction to a vector propagator on the charged black hole geometry. We find that the current dissipation is as efficient as possible and the transport lifetime coincides with the single-particle lifetime. In particular, in the case where the spectral density is that of a marginal Fermi liquid, the resistivity is linear in temperature.

preprint2013arXiv

Entanglement growth during thermalization in holographic systems

We derive in detail several universal features in the time evolution of entanglement entropy and other nonlocal observables in quenched holographic systems. The quenches are such that a spatially uniform density of energy is injected at an instant in time, exciting a strongly coupled CFT which eventually equilibrates. Such quench processes are described on the gravity side by the gravitational collapse of a thin shell that results in a black hole. Various nonlocal observables have a unified description in terms of the area of extremal surfaces of different dimensions. In the large distance limit, the evolution of an extremal surface, and thus the corresponding boundary observable, is controlled by the geometry around and inside the event horizon of the black hole, allowing us to identify regimes of pre-local- equilibration quadratic growth, post-local-equilibration linear growth, a memory loss regime, and a saturation regime with behavior resembling those in phase transitions. We also discuss possible bounds on the maximal rate of entanglement growth in relativistic systems.

preprint2013arXiv

Entanglement Tsunami: Universal Scaling in Holographic Thermalization

We consider the time evolution of entanglement entropy after a global quench in a strongly coupled holographic system, whose subsequent equilibration is described in the gravity dual by the gravitational collapse of a thin shell of matter resulting in a black hole. In the limit of large regions of entanglement, the evolution of entanglement entropy is controlled by the geometry around and inside the event horizon of the black hole, resulting in regimes of pre-local- equilibration quadratic growth (in time), post-local-equilibration linear growth, a late-time regime in which the evolution does not carry any memory of the size and shape of the entangled region, and a saturation regime with critical behavior resembling those in continuous phase transitions. Collectively, these regimes suggest a picture of entanglement growth in which an "entanglement tsunami" carries entanglement inward from the boundary. We also make a conjecture on the maximal rate of entanglement growth in relativistic systems.

preprint2013arXiv

GALAMOST: GPU-accelerated large-scale molecular simulation toolkit

A new molecular simulation toolkit composed of some lately developed force fields and specified models is presented to study the self-assembly, phase transition, and other properties of polymeric systems at mesoscopic scale by utilizing the computational power of GPUs. In addition, the hierarchical self-assembly of soft anisotropic particles and the problems related to polymerization can be studied by corresponding models included in this toolkit.

preprint2013arXiv

Holographic turbulence

We construct turbulent black holes in asymptotically AdS_4 spacetime by numerically solving Einstein equations. Both the dual holographic fluid and bulk geometry display signatures of an inverse cascade with the bulk geometry being well approximated by the fluid/gravity gradient expansion. We argue that statistically steady-state black holes dual to d dimensional turbulent flows have horizons which are approximately fractal with fractal dimension D=d+4/3.

preprint2013arXiv

Manipulating the magnetic anisotropy of cobalt doped titanium dioxide by carrier accumulation

Based on first-principles calculations, we predict that the magnetic anisotropy energy (MAE) of Co-doped TiO$_2$ sensitively depends on carrier accumulation. This magnetoelectric phenomenon provides a promising route to directly manipulate the magnetization direction of diluted magnetic semiconductor by external electric-fields. We calculate the band structures and reveal the origin of carrier-dependent MAE in k-space. In fact, the carrier accumulation shifts the Fermi energy and regulates the competing contributions to MAE. The first-principles calculations provide a straightforward way to design spintronics materials with electrically controllable spin direction.

preprint2013arXiv

Multicolor Ramsey numbers for triple systems

Given an $r$-uniform hypergraph $H$, the multicolor Ramsey number $r_k(H)$ is the minimum $n$ such that every $k$-coloring of the edges of the complete $r$-uniform hypergraph $K_n^r$ yields a monochromatic copy of $H$. We investigate $r_k(H)$ when $k$ grows and $H$ is fixed. For nontrivial 3-uniform hypergraphs $H$, the function $r_k(H)$ ranges from $\sqrt{6k}(1+o(1))$ to double exponential in $k$. We observe that $r_k(H)$ is polynomial in $k$ when $H$ is $r$-partite and at least single-exponential in $k$ otherwise. Erdős, Hajnal and Rado gave bounds for large cliques $K_s^r$ with $s\ge s_0(r)$, showing its correct exponential tower growth. We give a proof for cliques of all sizes, $s>r$, using a slight modification of the celebrated stepping-up lemma of Erdős and Hajnal. For 3-uniform hypergraphs, we give an infinite family with sub-double-exponential upper bound and show connections between graph and hypergraph Ramsey numbers. Specifically, we prove that $$r_k(K_3)\le r_{4k}(K_4^3-e)\le r_{4k}(K_3)+1,$$ where $K_4^3-e$ is obtained from $K_4^3$ by deleting an edge. We provide some other bounds, including single-exponential bounds for $F_5=\{abe,abd,cde\}$ as well as asymptotic or exact values of $r_k(H)$ when $H$ is the bow $\{abc,ade\}$, kite $\{abc,abd\}$, tight path $\{abc,bcd,cde\}$ or the windmill $\{abc,bde,cef,bce\}$. We also determine many new "small" Ramsey numbers and show their relations to designs. For example, the lower bound for $r_6(kite)=8$ is demonstrated by decomposing the triples of $[7]$ into six partial STS (two of them are Fano planes).

preprint2013arXiv

Probing renormalization group flows using entanglement entropy

In this paper we continue the study of renormalized entanglement entropy introduced in [1]. In particular, we investigate its behavior near an IR fixed point using holographic duality. We develop techniques which, for any static holographic geometry, enable us to extract the large radius expansion of the entanglement entropy for a spherical region. We show that for both a sphere and a strip, the approach of the renormalized entanglement entropy to the IR fixed point value contains a contribution that depends on the whole RG trajectory. Such a contribution is dominant, when the leading irrelevant operator is sufficiently irrelevant. For a spherical region such terms can be anticipated from a geometric expansion, while for a strip whether these terms have geometric origins remains to be seen.

preprint2013arXiv

Spontaneous Generation of Angular Momentum in Holographic Theories

The Schwarzschild black two-brane in four-dimensional anti-de Sitter space is dual to a finite temperature state in three-dimensional conformal field theory. We show that the solution acquires a nonzero angular momentum density when a gravitational Chern-Simons coupling is turned on in the bulk, even though the solution is not modified. A similar phenomenon is found for the Reissner-Nordstrom black two-brane with axionic coupling to the gauge field. We discuss interpretation of this phenomenon from the point of view of the boundary three-dimensional conformal field theory.

preprint2013arXiv

The biased odd cycle game

In this paper we consider biased Maker-Breaker games played on the edge set of a given graph $G$. We prove that for every $δ>0$ and large enough $n$, there exists a constant $k$ for which if $δ(G)\geq δn$ and $χ(G)\geq k$, then Maker can build an odd cycle in the $(1:b)$ game for $b=O(\frac{n}{\log^2 n})$. We also consider the analogous game where Maker and Breaker claim vertices instead of edges. This is a special case of the following well known and notoriously difficult problem due to Duffus, Łuczak and Rödl: is it true that for any positive constants $t$ and $b$, there exists an integer $k$ such that for every graph $G$, if $χ(G)\geq k$, then Maker can build a graph which is not $t$-colorable, in the $(1:b)$ Maker-Breaker game played on the vertices of $G$?

preprint2012arXiv

Ab initio calculation of the local magnetic moment in titanium doped zinc oxide with a corrected-band-gap scheme

The local magnetic moment of Ti:ZnO is calculated from first principles by using the corrected-band-gap scheme (CBGS). The results shows that the system is magnetic with the magnetization of 0.699 $μ_B$ per dopant. The origin of the local magnetic moment is considered to be the impurity band partially occupied by the donor electrons in the conduction band. Further, the impacts of applying Hubbard U to Ti-d orbital on the magnetic moment have been investigated.

preprint2012arXiv

Extremal graphs for blow-ups of cycles and trees

The \emph{blow-up} of a graph $H$ is the graph obtained from replacing each edge in $H$ by a clique of the same size where the new vertices of the cliques are all different. Erdős et al. and Chen et al. determined the extremal number of blow-ups of stars. Glebov determined the extremal number and found all extremal graphs for blow-ups of paths. We determined the extremal number and found the extremal graphs for the blow-ups of cycles and a large class of trees, when $n$ is sufficiently large. This generalizes their results. The additional aim of our note is to draw attention to a powerful tool, a classical decomposition theorem of Simonovits.

preprint2012arXiv

Gauge/String Duality, Hot QCD and Heavy Ion Collisions

Over the last decade, both experimental and theoretical advances have brought the need for strong coupling techniques in the analysis of deconfined QCD matter and heavy ion collisions to the forefront. As a consequence, a fruitful interplay has developed between analyses of strongly-coupled non-abelian plasmas via the gauge/string duality (also referred to as the AdS/CFT correspondence) and the phenomenology of heavy ion collisions. We review some of the main insights gained from this interplay to date. To establish a common language, we start with an introduction to heavy ion phenomenology and finite-temperature QCD, and a corresponding introduction to important concepts and techniques in the gauge/string duality. These introductory sections are written for nonspecialists, with the goal of bringing readers ranging from beginning graduate students to experienced practitioners of either QCD or gauge/string duality to the point that they understand enough about both fields that they can then appreciate their interplay in all appropriate contexts. We then review the current state-of-the art in the application of the duality to the description of the dynamics of strongly coupled plasmas, with emphases that include: its thermodynamic, hydrodynamic and transport properties; the way it both modifies the dynamics of, and is perturbed by, high-energy or heavy quarks passing through it; and the physics of quarkonium mesons within it. We seek throughout to stress the lessons that can be extracted from these computations for heavy ion physics as well as to discuss future directions and open problems for the field.

preprint2012arXiv

Holographic Vortex Liquids and Superfluid Turbulence

Superfluid turbulence, often referred to as quantum turbulence, is a fascinating phenomenon for which a satisfactory theoretical framework is lacking. Holographic duality provides a systematic new approach to studying quantum turbulence by mapping the dynamics of certain quantum theories onto the dynamics of classical gravity. We use this gravitational description to numerically construct turbulent flows in a holographic superfluid in two spatial dimensions. We find that the superfluid kinetic energy spectrum obeys the Kolmogorov -5/3 scaling law, as it does for turbulent flows in normal fluids. We trace this scaling to a direct energy cascade by injecting energy at long wavelengths and watching it flow to a short-distance scale set by the vortex core size, where dissipation by vortex annihilation and vortex drag becomes efficient. This is in sharp contrast with the inverse energy cascade of normal fluid turbulence in two dimensions. We also demonstrate that the microscopic dissipation spectrum has a simple geometric interpretation.

preprint2012arXiv

Luttinger's theorem, superfluid vortices, and holography

Strongly coupled field theories with gravity duals can be placed at finite density in two ways: electric field flux emanating from behind a horizon, or bulk charged fields outside of the horizon that explicitly source the density. We discuss field-theoretical observables that are sensitive to this distinction. If the charged fields are fermionic, we discuss a modified Luttinger's theorem that holds for holographic systems, in which the sum of boundary theory Fermi surfaces counts only the charge outside of the horizon. If the charged fields are bosonic, we show that the the resulting superfluid phase may be characterized by the coefficient of the transverse Magnus force on a moving superfluid vortex, which again is sensitive only to the charge outside of the horizon. For holographic systems these observables provide a field-theoretical way to distinguish how much charge is held by a dual horizon, but they may be useful in more general contexts as measures of deconfined (i.e. "fractionalized") charge degrees of freedom.

preprint2012arXiv

Momentum Broadening in Weakly Coupled Quark-Gluon Plasma (with a view to finding the quasiparticles within liquid quark-gluon plasma)

We calculate P(k_\perp), the probability distribution for an energetic parton that propagates for a distance L through a medium without radiating to pick up transverse momentum k_\perp, for a medium consisting of weakly coupled quark-gluon plasma. We use full or HTL self-energies in appropriate regimes, resumming each in order to find the leading large-L behavior. The jet quenching parameter \hat q is the second moment of P(k_\perp), and we compare our results to other determinations of this quantity in the literature, although we emphasize the importance of looking at P(k_\perp) in its entirety. We compare our results for P(k_\perp) in weakly coupled quark-gluon plasma to expectations from holographic calculations that assume a plasma that is strongly coupled at all length scales. We find that the shape of P(k_\perp) at modest k_\perp may not be very different in weakly coupled and strongly coupled plasmas, but we find that P(k_\perp) must be parametrically larger in a weakly coupled plasma than in a strongly coupled plasma at large enough k_\perp. This means that by looking for rare (but not exponentially rare) large-angle deflections of the jet resulting from a parton produced initially back-to-back with a hard photon, experimentalists can find the weakly coupled short-distance quark and gluon quasiparticles within the strongly coupled liquid quark-gluon plasma produced in heavy ion collisions, much as Rutherford found nuclei within atoms or Friedman, Kendall and Taylor found quarks within nucleons.

preprint2012arXiv

On coloring of fractional powers of graphs

For $m, n\in \N$, the fractional power $\Gmn$ of a graph $G$ is the $m$th power of the $n$-subdivision of $G$, where the $n$-subdivision is obtained by replacing each edge in $G$ with a path of length $n$. It was conjectured by Iradmusa that if $G$ is a connected graph with $Δ(G)\ge 3$ and $1<m<n$, then $χ(\Gmn)=ω(\Gmn)$. Here we show that the conjecture does not hold in full generality by presenting a graph $H$ for which $χ(H^{3/5})>ω(H^{3/5})$. However, we prove that the conjecture is true if $m$ is even. We also study the case when $m$ is odd, obtaining a general upper bound $χ(\Gmn)\leq ω(\Gmn)+2$ for graphs with $Δ(G)\geq 4$.

preprint2012arXiv

On the Turan number of forests

The Turan number of a graph H, ex(n,H), is the maximum number of edges in a graph on n vertices which does not have H as a subgraph. We determine the Turan number and find the unique extremal graph for forests consisting of paths when n is sufficiently large. This generalizes a result of Bushaw and Kettle [ Combinatorics, Probability and Computing 20:837--853, 2011]. We also determine the Turan number and extremal graphs for forests consisting of stars of arbitrary order.

preprint2012arXiv

Upper bounds on the size of 4- and 6-cycle-free subgraphs of the hypercube

In this paper we modify slightly Razborov's flag algebra machinery to be suitable for the hypercube. We use this modified method to show that the maximum number of edges of a 4-cycle-free subgraph of the n-dimensional hypercube is at most 0.6068 times the number of its edges. We also improve the upper bound on the number of edges for 6-cycle-free subgraphs of the n-dimensional hypercube from the square root of 2 - 1 to 0.3755 times the number of its edges. Additionally, we show that if the n-dimensional hypercube is considered as a poset, then the maximum vertex density of three middle layers in an induced subgraph without 4-cycles is at most 2.15121 times n choose n/2.

preprint2011arXiv

Dominating Sets in Triangulations on Surfaces

A dominating set D of a graph G is a set such that each vertex v of G is either in the set or adjacent to a vertex in the set. Matheson and Tarjan (1996) proved that any n-vertex plane triangulation has a dominating set of size at most n/3, and conjectured a bound of n/4 for n sufficiently large. King and Pelsmajer recently proved this for graphs with maximum degree at most 6. Plummer and Zha (2009) and Honjo, Kawarabayashi, and Nakamoto (2009) extended the n/3 bound to triangulations on surfaces. We prove two related results: (i) There is a constant c such that any n-vertex plane triangulation with maximum degree at most 6 has a dominating set of size at most n/6 + c. (ii) For any surface S, nonnegative t, and epsilon > 0, there exists C such that for any n-vertex triangulation on S with at most t vertices of degree other than 6, there is a dominating set of size at most n(1/6 + epsilon) + C. As part of the proof, we also show that any n-vertex triangulation of a non-orientable surface has a non-contractible cycle of length at most 2sqrt(n). Albertson and Hutchinson (1986) proved that for n-vertex triangulation of an orientable surface other than a sphere has a non-contractible cycle of length sqrt(2n), but no similar result was known for non-orientable surfaces.

preprint2011arXiv

Emergent quantum criticality, Fermi surfaces, and AdS2

Gravity solutions dual to d-dimensional field theories at finite charge density have a near-horizon region which is AdS_2 x R^{d-1}. The scale invariance of the AdS_2 region implies that at low energies the dual field theory exhibits emergent quantum critical behavior controlled by a (0+1)-dimensional CFT. This interpretation sheds light on recently-discovered holographic descriptions of Fermi surfaces, allowing an analytic understanding of their low-energy excitations. For example, the scaling behavior near the Fermi surfaces is determined by conformal dimensions in the emergent IR CFT. In particular, when the operator is marginal in the IR CFT, the corresponding spectral function is precisely of the "Marginal Fermi Liquid" form, postulated to describe the optimally doped cuprates.

preprint2011arXiv

Holographic non-Fermi liquid fixed points

Techniques arising from string theory can be used to study assemblies of strongly-interacting fermions. Via this `holographic duality', various strongly-coupled many body systems are solved using an auxiliary theory of gravity. Simple holographic realizations of finite density exhibit single-particle spectral functions with sharp Fermi surfaces, of a form distinct from those of the Landau theory. The self-energy is given by a correlation function in an infrared fixed point theory which is represented by an AdS_2 region in the dual gravitational description. Here we describe in detail the gravity calculation of this IR correlation function. This article is a contribution to a special issue of Phil. Trans. A on the normal state of the cuprates; as such, we also provide some review and context.

preprint2011arXiv

Impact of heterogenous prior beliefs and disclosed insider trades

In this paper, we present a multi-period trading model by assuming that traders face not only asymmetric information but also heterogenous prior beliefs, under the requirement that the insider publicly disclose his stock trades after the fact. We show that there is an equilibrium in which the irrational insider camouflages his trades with a noise component so that his private information is revealed slowly and linearly whenever he is overconfident or underconfident. We also investigate the relationship between the heterogeneous beliefs and the trade intensity in the presence of trade disclosure, and show that the weights on asymmetric information and heterogeneous prior beliefs are opposite in sign and they change alternatively in the next period. Under the requirement of disclosure, the irrational insider trades more aggressively and leads to smaller market depth. Moreover, the co-existence of "public disclosure requirement" and "heterogeneous prior beliefs" leads to the fluctuant multi-period expected profits and a larger total expected trading volume which is positively related to the degree of heterogeneity. More importantly, even public disclosure may lead to negative profits of the irrational insider's in some periods, inside trading remains profitable from the whole trading period.

preprint2011arXiv

Inside Trading, Public Disclosure and Imperfect Competition

In this paper, we present a multi-period trading model in the style of Kyle (1985)'s inside trading model, by assuming that there are at least two insiders in the market with long-lived private information, under the requirement that each insider publicly discloses his stock trades after the fact. Based on this model, we study the influences of "public disclosure" and "competition among insiders" on the trading behaviors of insiders. We find that the "competition among insiders" leads to higher effective price and lower insiders' profits, and the "public disclosure" makes each insider play a mixed strategy in every round except the last one. An interesting find is that as the total number of auctions goes to infinity, the market depth and the trading intensity at the first auction are all constants with the requirement of "public disclosure", while the market depth at the first auction goes to zero and the trading intensity of the first period goes to infinity without the requirement of "public disclosure".Moreover, we give the exact speed of the revelation of the private information, and show that all information is revealed immediately and the market depth goes to infinity immediately as trading happens infinitely frequently.

preprint2011arXiv

Integrating out geometry: Holographic Wilsonian RG and the membrane paradigm

We formulate a holographic Wilsonian renormalization group flow for strongly coupled systems with a gravity dual, motivated by the need to extract efficiently low energy behavior of such systems. Starting with field theories defined on a cut-off surface in a bulk spacetime, we propose that integrating out high energy modes in the field theory should correspond to integrating out a part of the bulk geometry. We describe how to carry out this procedure in practice in the classical gravity approximation using examples of scalar and vector fields. By integrating out bulk degrees of freedom all the way to a black hole horizon, this formulation defines a refined version of the black hole membrane paradigm. Furthermore, it also provides a derivation of the semi-holographic description of low energy physics.

preprint2011arXiv

Momentum Broadening in Weakly Coupled Quark-Gluon Plasma

We calculate the probability distribution P(k_\perp) for the momentum perpendicular to its original direction of motion that an energetic quark or gluon picks up as it propagates through weakly coupled quark-gluon plasma in thermal equilibrium.

preprint2011arXiv

Semi-local quantum liquids

Gauge/gravity duality applied to strongly interacting systems at finite density predicts a universal intermediate energy phase to which we refer as a semi-local quantum liquid. Such a phase is characterized by a finite spatial correlation length, but an infinite correlation time and associated nontrivial scaling behavior in the time direction, as well as a nonzero entropy density. For a holographic system at a nonzero chemical potential, this unstable phase sets in at an energy scale of order of the chemical potential, and orders at lower energies into other phases; examples include superconductors and antiferromagnetic-type states. In this paper we give examples in which it also orders into Fermi liquids of "heavy" fermions. While the precise nature of the lower energy state depends on the specific dynamics of the individual system, we argue that the semi-local quantum liquid emerges universally at intermediate energies through deconfinement (or equivalently fractionalization). We also discuss the possible relevance of such a semi-local quantum liquid to heavy electron systems and the strange metal phase of high temperature cuprate superconductors.

preprint2011arXiv

Synchrotron radiation in strongly coupled conformal field theories

Using gauge/gravity duality, we compute the energy density and angular distribution of the power radiated by a quark undergoing circular motion in strongly coupled ${\cal N}=4$ supersymmetric Yang-Mills (SYM) theory. We compare the strong coupling results to those at weak coupling, and find the same angular distribution of radiated power, up to an overall prefactor. In both regimes, the angular distribution is in fact similar to that of synchrotron radiation produced by an electron in circular motion in classical electrodynamics: the quark emits radiation in a narrow beam along its velocity vector with a characteristic opening angle $α\sim 1/γ$. To an observer far away from the quark, the emitted radiation appears as a short periodic burst, just like the light from a lighthouse does to a ship at sea. Our strong coupling results are valid for any strongly coupled conformal field theory with a dual classical gravity description.

preprint2011arXiv

Transverse Momentum Broadening and the Jet Quenching Parameter, Redux

We use Soft Collinear Effective Theory (SCET) to analyze the transverse momentum broadening, or diffusion in transverse momentum space, of an energetic parton propagating through quark-gluon plasma. Since we neglect the radiation of gluons from the energetic parton, we can only discuss momentum broadening, not parton energy loss. The interaction responsible for momentum broadening in the absence of radiation is that between the energetic (collinear) parton and the Glauber modes of the gluon fields in the medium. We derive the effective Lagrangian for this interaction, and we show that the probability for picking up transverse momentum k_\perp is given by the Fourier transform of the expectation value of two transversely separated light-like path-ordered Wilson lines. This yields a field theoretical definition of the jet quenching parameter \hat q, and shows that this can be interpreted as a diffusion constant. We close by revisiting the calculation of \hat q for the strongly coupled plasma of N=4 SYM theory, showing that previous calculations need some modifications that make them more straightforward and do not change the result.

preprint2010arXiv

A controlled expansion for certain non-Fermi liquid metals

The destruction of Fermi liquid behavior when a gapless Fermi surface is coupled to a fluctuating gapless boson field is studied theoretically. This problem arises in a number of different contexts in quantum many body physics. Examples include fermions coupled to a fluctuating transverse gauge field pertinent to quantum spin liquid Mott insulators, and quantum critical metals near a Pomeranchuk transition. We develop a new controlled theoretical approach to determining the low energy physics. Our approach relies on combining an expansion in the inverse number (N) of fermion species with a further expansion in the parameter ε= z_b -2 where z_b is the dynamical critical exponent of the boson field. We show how this limit allows a systematic calculation of the universal low energy physics of these problems. The method is illustrated by studying spinon fermi surface spin liquids, and a quantum critical metal at a second order electronic nematic phase transition. We calculate the low energy single particle spectra, and various interesting two particle correlation functions. In some cases deviations from the popular Random Phase Approximation results are found. Some of the same universal singularities are also calculated to leading non-vanishing order using a perturbative renormalization group calculation at small N extending previous results of Nayak and Wilczek. Implications for quantum spin liquids, and for Pomeranchuk transitions are discussed. For quantum critical metals at a nematic transition we show that the tunneling density of states has a power law suppression at low energies.

preprint2010arXiv

From black holes to strange metals

Since the mid-eighties there has been an accumulation of metallic materials whose thermodynamic and transport properties differ significantly from those predicted by Fermi liquid theory. Examples of these so-called non-Fermi liquids include the strange metal phase of high transition temperature cuprates, and heavy fermion systems near a quantum phase transition. We report on a class of non-Fermi liquids discovered using gauge/gravity duality. The low energy behavior of these non-Fermi liquids is shown to be governed by a nontrivial infrared (IR) fixed point which exhibits nonanalytic scaling behavior only in the temporal direction. Within this class we find examples whose single-particle spectral function and transport behavior resemble those of strange metals. In particular, the contribution from the Fermi surface to the conductivity is inversely proportional to the temperature. In our treatment these properties can be understood as being controlled by the scaling dimension of the fermion operator in the emergent IR fixed point.

preprint2010arXiv

Jet Quenching Parameter via Soft Collinear Effective Theory (SCET)

We analyze the transverse momentum broadening in the absence of radiation of an energetic parton propagating through quark-gluon plasma via Soft Collinear Effective Theory (SCET). We show that the probability for picking up transverse momentum k_\perp is given by the Fourier transform of the expectation value of two transversely separated light-like path-ordered Wilson lines. The subtleties about the ordering of operators do not change the \hat q value for the strongly coupled plasma of N=4 SYM theory.

preprint2010arXiv

Quantum phase transitions in holographic models of magnetism and superconductors

We study a holographic model realizing an "antiferromagnetic" phase in which a global SU(2) symmetry representing spin is broken down to a U(1) by the presence of a finite electric charge density. This involves the condensation of a neutral scalar field in a charged AdS black hole. We observe that the phase transition for both neutral and charged (as in the standard holographic superconductor) order parameters can be driven to zero temperature by a tuning of the UV conformal dimension of the order parameter, resulting in a quantum phase transition of the Berezinskii-Kosterlitz-Thouless type. We also characterize the antiferromagnetic phase and an externally forced ferromagnetic phase by showing that they contain the expected spin waves with linear and quadratic dispersions respectively.

preprint2010arXiv

Radiation of a circulating quark in strongly coupled N=4 super Yang-Mills theory

The energy density and angular distribution of power radiated by a quark undergoing circular motion in strongly coupled ${\cal N}=4$ supersymmetric Yang-Mills (SYM) theory is computed using gauge/gravity duality. The results are qualitatively similar to that of synchrotron radiation produced by an electron in circular motion in classical electrodynamics: At large velocities the quark emits radiation in a narrow beam along its velocity vector with a characteristic opening angle $α\sim 1/γ$ and radial thickness scaling like $\sim 1/γ^3$.

preprint2009arXiv

Non-Fermi liquids from holography

We report on a potentially new class of non-Fermi liquids in (2+1)-dimensions. They are identified via the response functions of composite fermionic operators in a class of strongly interacting quantum field theories at finite density, computed using the AdS/CFT correspondence. We find strong evidence of Fermi surfaces: gapless fermionic excitations at discrete shells in momentum space. The spectral weight exhibits novel phenomena, including particle-hole asymmetry, discrete scale invariance, and scaling behavior consistent with that of a critical Fermi surface postulated by Senthil.

preprint2008arXiv

A Bohr-Sommerfeld quantization formula for quasinormal frequencies of AdS black holes

We derive a quantization formula of Bohr-Sommerfeld type for computing quasinormal frequencies for scalar perturbations in an AdS black hole in the limit of large scalar mass or spatial momentum. We then apply the formula to find poles in retarded Green functions of boundary CFTs on $R^{1,d-1}$ and $RxS^{d-1}$. We find that when the boundary theory is perturbed by an operator of dimension $Δ>> 1$, the relaxation time back to equilibrium is given at zero momentum by ${1 \over ΔπT} << {1 \over πT}$. Turning on a large spatial momentum can significantly increase it. For a generic scalar operator in a CFT on $R^{1,d-1}$, there exists a sequence of poles near the lightcone whose imaginary part scales with momentum as $p^{-{d-2 \over d+2}}$ in the large momentum limit. For a CFT on a sphere $S^{d-1}$ we show that the theory possesses a large number of long-lived quasiparticles whose imaginary part is exponentially small in momentum.

preprint2008arXiv

Stirring Strongly Coupled Plasma

We determine the energy it takes to move a test quark along a circle of radius L with angular frequency w through the strongly coupled plasma of N=4 supersymmetric Yang-Mills (SYM) theory. We find that for most values of L and w the energy deposited by stirring the plasma in this way is governed either by the drag force acting on a test quark moving through the plasma in a straight line with speed v=Lw or by the energy radiated by a quark in circular motion in the absence of any plasma, whichever is larger. There is a continuous crossover from the drag-dominated regime to the radiation-dominated regime. In the crossover regime we find evidence for significant destructive interference between energy loss due to drag and that due to radiation as if in vacuum. The rotating quark thus serves as a model system in which the relative strength of, and interplay between, two different mechanisms of parton energy loss is accessible via a controlled classical gravity calculation. We close by speculating on the implications of our results for a quark that is moving through the plasma in a straight line while decelerating, although in this case the classical calculation breaks down at the same value of the deceleration at which the radiation-dominated regime sets in.

preprint2008arXiv

Viscosity Bound Violation in Higher Derivative Gravity

Motivated by the vast string landscape, we consider the shear viscosity to entropy density ratio in conformal field theories dual to Einstein gravity with curvature square corrections. After field redefinitions these theories reduce to Gauss-Bonnet gravity, which has special properties that allow us to compute the shear viscosity nonperturbatively in the Gauss-Bonnet coupling. By tuning of the coupling, the value of the shear viscosity to entropy density ratio can be adjusted to any positive value from infinity down to zero, thus violating the conjectured viscosity bound. At linear order in the coupling, we also check consistency of four different methods to calculate the shear viscosity, and we find that all of them agree. We search for possible pathologies associated with this class of theories violating the viscosity bound.

preprint2007arXiv

Closed strings from decaying D-branes

We compute the emission of closed string radiation from homogeneous rolling tachyons. For an unstable decaying D$p$-brane the radiated energy is infinite to leading order for $p\leq 2$ and finite for $p>2$. The closed string state produced by a decaying brane is closely related to the state produced by D-instantons at a critical Euclidean distance from $t=0$. In the case of a D0 brane one can cutoff this divergence so that we get a finite energy final state which would be the state that the brane decays into.

preprint2007arXiv

Hagedorn divergences and tachyon potential

We consider the critical behavior for a string theory near the Hagedorn temperature. We use the factorization of the worldsheet to isolate the Hagedorn divergences at all genera. We show that the Hagedorn divergences can be resummed by introducing double scaling limits, which smooth the divergences. The double scaling limits also allow one to extract the effective potential for the thermal scalar. For a string theory in an asymptotic anti-de Sitter (AdS) spacetime, the AdS/CFT correspondence implies that the critical Hagedorn behavior and the relation with the effective potential should also arise from the boundary Yang-Mills theory. We show that this is indeed the case. In particular we find that the free energy of a Yang-Mills theory contains ``vortex'' contributions at finite temperature. Yang-Mills Feynman diagrams with vortices can be identified with contributions from boundaries of moduli space on the string theory side.

preprint2007arXiv

Wilson loops in heavy ion collisions and their calculation in AdS/CFT

Expectation values of Wilson loops define the nonperturbative properties of the hot medium produced in heavy ion collisions that arise in the analysis of both radiative parton energy loss and quarkonium suppression. We use the AdS/CFT correspondence to calculate the expectation values of such Wilson loops in the strongly coupled plasma of N=4 super Yang-Mills (SYM) theory, allowing for the possibility that the plasma may be moving with some collective flow velocity as is the case in heavy ion collisions. We obtain the N=4 SYM values of the jet quenching parameter $\hat q$, which describes the energy loss of a hard parton in QCD, and of the velocity-dependence of the quark-antiquark screening length for a moving dipole as a function of the angle between its velocity and its orientation. We show that if the quark-gluon plasma is flowing with velocity v_f at an angle theta with respect to the trajectory of a hard parton, the jet quenching parameter $\hat q$ is modified by a factor gamma_f(1-v_f cos theta), and show that this result applies in QCD as in N=4 SYM. We discuss the relevance of the lessons we are learning from all these calculations to heavy ion collisions at RHIC and at the LHC. Furthermore, we discuss the relation between our results and those obtained in other theories with gravity duals, showing in particular that the ratio between $\hat q$ in any two conformal theories with gravity duals is the square root of the ratio of their central charges. This leads us to conjecture that in nonconformal theories $\hat q$ defines a quantity that always decreases along renormalization group trajectories and allows us to use our calculation of $\hat q$ in N=4 SYM to make a conjecture for its value in QCD.

preprint2006arXiv

Bulk-cone singularities & signatures of horizon formation in AdS/CFT

We discuss the relation between singularities of correlation functions and causal properties of the bulk spacetime in the context of the AdS/CFT correspondence. In particular, we argue that the boundary field theory correlation functions are singular when the insertion points are connected causally by a bulk null geodesic. This implies the existence of "bulk-cone singularities" in boundary theory correlation functions which lie inside the boundary light-cone. We exhibit the pattern of singularities in various asymptotically AdS spacetimes and argue that this pattern can be used to probe the bulk geometry. We apply this correspondence to the specific case of shell collapse in AdS/CFT and indicate a sharp feature in the boundary observables corresponding to black hole event horizon formation.

preprint2005arXiv

Finite temperature effective action, AdS_5 black holes, and 1/N expansion

We propose a phenomenological matrix model to study string theory in AdS_5 \times S_5 in the canonical ensemble. The model reproduces all the known qualitative features of the theory. In particular, it gives a simple effective potential description of Euclidean black hole nucleation and the tunnelling between thermal AdS and the big black hole. It also has some interesting predictions. We find that there exists a critical temperature at which the Euclidean small black hole undergoes a Gross-Witten phase transition. We identify the phase transition with the Horowitz-Polchinski point where the black hole horizon size becomes comparable to the string scale. The appearance of the Hagedorn divergence of thermal AdS is due to the merger of saddle points corresponding to the Euclidean small black hole and thermal AdS. The merger can be described in terms of a cusp (A_3) catastrophe and divergences at the perturbative string level are smoothed out at finite string coupling using standard techniques of catastrophe theory.

preprint2001arXiv

*-Trek III: The Search for Ramond-Ramond Couplings

We give a detailed discussion of the disk amplitudes with one closed string insertion, which we used to construct the supergravity couplings of noncommutative D-branes to the RR potentials, given in hep-th/0104139. We prove the inclusion of Elliott's formula, the integer-valued modification of the noncommutative Chern character, to all orders in the gauge field. We also give a detailed comparison between the form of the result in which Elliott's formula is manifest, and the form expressed in Matrix model variables.

preprint1999arXiv

Scattering in Anti-de Sitter Space and Operator Product Expansion

We develop a formalism to evaluate generic scalar exchange diagrams in AdS_{d+1} relevant for the calculation of four-point functions in AdS/CFT correspondence. The result may be written as an infinite power series of functions of cross-ratios. Logarithmic singularities appear in all orders whenever the dimensions of involved operators satisfy certain relations. We show that the AdS_{d+1} amplitude can be written in a form recognisable as the conformal partial wave expansion of a four-point function in CFT_{d} and identify the spectrum of intermediate operators. We find that, in addition to the contribution of the scalar operator associated with the exchanged field in the AdS diagram, there are also contributions of some other operators which may possibly be identified with two-particle bound states in AdS. The CFT interpretation also provides a useful way to ``regularize'' the logarithms appearing in AdS amplitude.

preprint1996arXiv

Quantum Hair, Instantons, and Black Hole Thermodynamics: Some New Results

We present results obtained by a consideration of the non-classical energy momentum tensor associated with Euclidean Instantons outside the event horizon of black holes. We demonstrate here how this allows an analytic estimate to be made of the effect of discrete quantum hair on the temperature of the black hole, in which the role of violations of the weak energy condition associated with instantons is made explicit, and in which the previous results of Coleman, Preskill, and Wilczek are extended. Last, we demonstrate how the existence of a non-classical electric field outside the event horizon of black holes, uncovered by these authors, can be identified with a well-known effect in the Abelian-Higgs model in two dimensions. In this case, there is a one-to-one connection between the discrete charge of a black hole and a topological phase in two dimensions.

preprint1995arXiv

Perturbed Electroweak Strings and Fermion Zero Modes

The effect of perturbations of electroweak strings on quark zero modes is studied in 2+1 and 3+1 dimensions. As first discovered by Naculich, it is found that the bosonic perturbations that destabilize the string give a mass to the zero modes and also lift their degeneracy. The effect of the zero modes on the stability of the string is discussed qualitatively and we argue that the fermionic vacuum instability found by Naculich should lead to a distortion of the bosonic string but not be responsible for decay.

Hong Liu

What is connected

Connect this record

See the researcher in context

Building this map preview

133 published item(s)

MoPO: Incorporating Motion Prior for Occluded Human Mesh Recovery

A Gated Cross-domain Collaborative Network for Underwater Object Detection

A systematic formulation of chiral anomalous magnetohydrodynamics

Achieving Domain Generalization in Underwater Object Detection by Domain Mixup and Contrastive Learning

Many Hamiltonian subsets in large graphs with given density

Polynomial Schur's theorem

Advancing Semi-Supervised Task Oriented Dialog Systems by JSA Learning of Discrete Latent Variable Models

AO2-DETR: Arbitrary-Oriented Object Detection Transformer

Breast Cancer Molecular Subtypes Prediction on Pathological Images with Discriminative Patch Selecting and Multi-Instance Learning

Clique immersion in graphs without fixed bipartite graph

Contrastive Learning from Spatio-Temporal Mixed Skeleton Sequences for Self-Supervised Skeleton-Based Action Recognition

Disjoint isomorphic balanced clique subdivisions

Exploiting Temporal Contexts with Strided Transformer for 3D Human Pose Estimation

Exponential decay of intersection volume with applications on list-decodability and Gilbert-Varshamov type bound

Facial-Sketch Synthesis: A New Challenge

High-order Photonic Cavity Modes Enabled 3D Structural Color

How to build a pillar: a proof of Thomassen's conjecture

Identity-Sensitive Knowledge Propagation for Cloth-Changing Person Re-identification

Learning Deep Direct-Path Relative Transfer Function for Binaural Sound Source Localization

MHFormer: Multi-Hypothesis Transformer for 3D Human Pose Estimation

Mission Apollo: Landing Optical Circuit Switching at Datacenter Scale

Multi-Scale Spatial Temporal Graph Convolutional Network for Skeleton-Based Action Recognition

PetLock:A Genderless and Standard Interface for the Future On-orbit Construction

Prototype Design and Efficiency Analysis of a Novel Robot Drive Based on 3K-H-V Topology

Recent advances and clinical applications of deep learning in medical image analysis

Recent Advances in Tunable Metasurfaces: Materials, Design and Applications

Self-supervised Learning is More Robust to Dataset Imbalance

Shape of the asymptotic maximum sum-free sets in integer lattice grids

Simultaneous Alignment and Surface Regression Using Hybrid 2D-3D Networks for 3D Coherent Layer Segmentation of Retina OCT Images

Snowmass White Paper: Effective Field Theories for Condensed Matter Systems

Snowmass White Paper: New ideas for many-body quantum systems from string theory and black holes

SP-SEDT: Self-supervised Pre-training for Sound Event Detection Transformer

Spatiotemporal Propagation Learning for Network-Wide Flight Delay Prediction

SRP-DNN: Learning Direct-Path Phase Difference for Multiple Moving Sound Source Localization

Transformer for Single Image Super-Resolution

Transformers Improve Breast Cancer Diagnosis from Unregistered Multi-View Mammograms

Virtual Adversarial Training for Semi-supervised Breast Mass Classification

Weakly-supervised 3D Human Pose Estimation with Cross-view U-shaped Graph Convolutional Network

Bound entanglement in thermalized states and black hole radiation

Mixed-state entanglement and information recovery in thermalized states and evaporating black holes

On systems of maximal quantum chaos

Anti-Bandit Neural Architecture Search for Model Defense

Applying a random projection algorithm to optimize machine learning model for breast lesion classification

Bi-directional Exponential Angular Triplet Loss for RGB-Infrared Person Re-Identification

Dynamical Phase Transition from Nonequilibrium Dynamics of Dark Solitons

Guided learning for weakly-labeled semi-supervised sound event detection

Hadamard Matrix Guided Online Hashing

Multi-Branch Learning for Weakly-Labeled Sound Event Detection

Online Initialization and Extrinsic Spatial-Temporal Calibration for Monocular Visual-Inertial Odometry

Projection & Probability-Driven Black-Box Attack

Quantum many-body physics from a gravitational lens

Self-Refining Deep Symmetry Enhanced Network for Rain Removal

Spatial Pyramid Based Graph Reasoning for Semantic Segmentation

Specialized Decision Surface and Disentangled Feature for Weakly-Supervised Polyphonic Sound Event Detection

The exact minimum number of triangles in graphs of given order and size

Video Logo Retrieval based on local Features

When Dictionary Learning Meets Deep Learning: Deep Dictionary Learning and Coding Network for Image Recognition with Limited Data

WQT and DG-YOLO: towards domain generalization in underwater object detection

Compact all-fiber polarization-independent up-conversion single-photon detector

Sample phase gradient and fringe phase shift in dual phase grating X-ray interferometry

Tungsten Boride: a 2D Multiple Dirac Semimetal for Hydrogen Evolution Reaction

Coulomb drag in topological insulator films

Densities in large permutations and parameter testing

Flat Helical Nanosieves

Local conditions for exponentially many subdivisions

Ordinal Constrained Binary Code Learning for Nearest Neighbor Search

Orientation Driven Bag of Appearances for Person Re-identification

Spread of entanglement and causality

Supervised Matrix Factorization for Cross-Modality Hashing

The number of subsets of integers with no $k$-term arithmetic progression

Continuously Shaping Orbital Angular Momentum with an Analog Optical Vortex Transmitter

Defect formation beyond Kibble-Zurek mechanism and holography

Intersecting families of discrete structures are typically trivial

Off-shell hydrodynamics from holography